Delayed Pituitary Adenylate Cyclase–Activating Polypeptide Delivery After Brain Stroke Improves Functional Recovery by Inducing M2 Microglia/Macrophage Polarization
Background and Purpose—Until now, except thrombolysis, the therapeutical strategies targeting the acute phase of cerebral ischemia have been proven ineffective, and no approach is available to attenuate the delayed cell death mechanisms and the resulting functional deficits in the late phase. Then, we investigated whether a targeted and delayed delivery of pituitary adenylate cyclase–activating polypeptide (PACAP), a peptide known to exert neuroprotective activities, may dampen delayed pathophysiological processes improving functional recovery.
Methods—Three days after permanent focal ischemia, PACAP-producing stem cells were transplanted intracerebro ventricularly in nonimmunosuppressed mice. At 7 and 14 days post ischemia, the effects of this stem cell–based targeted delivery of PACAP on functional recovery, volume lesions, and inflammatory processes were analyzed.
Results—The delivery of PACAP in the vicinity of the infarct zone 3 days post stroke promotes fast, stable, and efficient functional recovery. This was correlated with a modulation of the postischemic inflammatory response. Transcriptomic and Ingenuity Pathway Analysis–based bioinformatic analyses identified several gene networks, functions, and key transcriptional factors, such as nuclear factor-κB, C/EBP-β, and Notch/RBP-J as PACAP’s potential targets. Such PACAP-dependent immunomodulation was further confirmed by morphometric and phenotypic analyses of microglial cells showing increased number of Arginase-1+ cells in mice treated with PACAP-expressing cells specifically, demonstrating the redirection of the microglial response toward a neuroprotective M2 phenotype.
Conclusions—Our results demonstrated that immunomodulatory strategies capable of redirecting the microglial response toward a neuroprotective M2 phenotype in the late phase of brain ischemia could represent attractive options for stroke treatment in a new and unexploited therapeutical window.
Stroke is a leading cause of death and long-term disabilities worldwide. Despite years of intense research and preclinical identification of numerous potential neuroprotective compounds, the only available treatment for brain ischemia relies on thrombolysis through injection of a recombinant tissue-type plasminogen activator. However, the treatment benefits to <10% of stroke victims because of a narrow therapeutical time window (<4.5 hours after stroke onset) and side effects.1 Consequently, there is a crucial need for the development of other strategies that could target later phases of the pathophysiological cascade of events after stroke.
Since its initial discovery, several studies have highlighted the neuroprotective effect of pituitary adenylate cyclase–activating polypeptide (PACAP)2 in in vitro and in vivo models of neurodegenerative diseases.3,4 Administered either before or few hours after middle cerebral artery occlusion, PACAP reduces the infarct volume area and improves functional outcomes.5–7 Beside its well-known antiapoptotic activity, the neuropeptide PACAP exerts potent anti-inflammatory properties on innate immune compartment as illustrated by the decrease of the production of proinflammatory mediators interleukin (IL)-12, tumor necrosis factor (TNF)-α, and nitric oxide and the induction of the anti-inflammatory cytokine IL-10 in PACAP-treated macrophages stimulated by lipopolysaccharides.8–10 Whether PACAP acts directly by reducing apoptotic neuronal death11 or indirectly via modulation of the inflammatory processes12,13 is not fully understood yet. Nevertheless, there is growing evidence that PACAP could be of therapeutic value for the treatment of disorders involving apoptosis and neuroinflammation. However, its clinical use is restricted by its short half-life in systemic circulation (5–10 minutes in human blood)4,14 and in the context of brain stroke, by the difficulties in accessing the damaged cerebral tissues especially after cerebrovascular disruption. To circumvent these limitations, we propose a strategy of drug delivery based on the ability of genetically modified stem cells to migrate and locally deliver PACAP in the vicinity of the infarct area after brain transplantation. A key aspect of the experimental system is based on the use of stem cells and recipient mice of identical genetic background to obviate the immunosuppressive treatments that could interfere with the mechanisms under study. In this model, PACAP acts as a potent regulator of the microglial response in vivo at delayed time points after the stroke onset, leading to efficient functional recovery.
Materials and Methods
Materials and Methods are detailed in the online-only Data Supplement. All animal procedures were conformed to the French recommendations for the care and use of laboratory animals and approved by the regional ethics committee (authorization number, N/01-07-09/15/07-12).
Twelve- to 16-week-old male mice (129Sv mice or Cx3cr1+/GFP transgenic mice) were anesthetized with isoflurane and submitted to focal permanent cerebral ischemia by electrocauterization of the right middle cerebral artery.15 Seventy-two hours after occlusion, ischemized mice were subjected to a stereotaxic intracerebroventricular injection of 5.104 embryonic stem (ES) or PACAP-expressing ES (ES-P) cells or saline solution according to experimental groups.
Post-traumatic neurological impairment was analyzed using a mouse neurological severity score, as previously described.16 The hole-board test permitted an evaluation of the exploratory behavior and motor coordination of mice.17
At 7 and 14 dpi, animals were deeply anesthetized with pentobarbital and euthanized by transcardiac perfusion with 4% paraformaldehyde. Brain coronal slices of 30-µm-thick were cut using a vibrating blade microtome (VT1200S, Leica). The ischemic area was determined using a thionin-based morphometric analysis and 3-dimensional (3D) reconstruction using Imaris Software (Bitplane).
Free-floating 30-µm-thick sections were incubated in blocking buffer followed by incubation with mouse liver Arginase-1 (Arg-1) antibody (Abcam, reference ab60176) revealed by secondary Alexa 568 conjugated antibody (Invitrogen, reference A11057). Sections were mounted and viewed under a confocal laser scanning microscope (TCS SP5, Leica).
Morphological Analysis of Microglial Cells
Morphometric measurements of cell shape were performed on green fluorescent protein (GFP)–expressing microglia of Cx3cr1+/GFP transgenic mice. Morphological parameters characterizing the microglial activation were determined using IMARIS software (Bitplane).
Gene Expression Analysis
At 7 and 14 dpi, total RNAs of mouse from contra- and ipsilateral hemispheres were extracted from a series of 40-µm-thick brain slices using TRI-reagent (Sigma) according to the manufacturer’s protocol. Total RNAs were reverse transcribed into single-stranded cDNA and TNF-α, IL-10, and Ym-1 encoding cDNA quantified by real-time polymerase chain reaction on an ABI Prism 7500 Sequence Detection System (Life Technologies) using GAPDH as a reference gene. For transcriptomic analysis, mRNAs extracted from ipsilateral hemispheres of saline- and ES and ES-P cell–injected mice (n=3 for each group) were reverse transcribed into cDNA and loaded on a TaqMan® OpenArray® Mouse Inflammation Panel plate (Life Technologies, reference 4475393). Real-time polymerase chain reaction and quantitative gene expression were performed on a QuantStudio 12K Flex Real-Time Polymerase Chain Reaction System (Life Technologies).
The transcriptional signatures obtained from TaqMan® OpenArray® Gene Expression analysis were further analyzed through Ingenuity Pathway Analysis (IPA) software tools.
All values were expressed as mean±SEM. Statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software Inc.) and based on 1-way ANOVA followed by post hoc Tukey test for multiple comparisons. Statistical significance was set at P<0.05.
Delayed and Targeted Delivery of PACAP Promotes Functional Recovery After Stroke
To study the functional recovery resulting specifically from delayed and targeted PACAP delivery, wild-type ES cells and ES-P cells, producing and releasing PACAP as demonstrated at the transcriptomic level by real-time polymerase chain reaction and at the protein level by Western blot, immunocytofluorescence, and radioimmunoassay (Figure I in the online-only Data Supplement), were injected intracerebroventricularly 3 days after permanent middle cerebral artery occlusion in mice. The functional deficits were assessed at 7 and 14 days post ischemia (dpi) using neurological severity scores and hole-board tests. Although the delayed injection of ES cells did not reduce the functional deficits after stroke, the graft of ES-P cells promoted fast and stable functional recovery as illustrated by the neurological severity score at 7 and 14 dpi when compared with saline and ES groups (Figure 1A, upper). Similarly, motor coordination, typically affected after brain ischemia, was restored by ES-P cell injection in contrast to ES-grafted cells to levels comparable with those in sham-operated animals as measured by stumble frequency in hole-board tests (Figure 1A, middle), whereas there were no differences in ambulatory times among the 4 experimental groups (Figure 1A, lower).
Improvement of Functional Recovery Is Not Associated With a Reduction of the Lesion Size
Because the reduction of functional deficits could result from a decrease of tissue damages or cell replacement after stem cell transplantation, we measured the infarct and edema volumes using thionin-based morphometric analysis (Figure 1B–1D). Volumetric analyses of the ischemic lesions in the fronto-parietal cortex (Figure 1B) failed to reveal differences in lesion size (P>0.05; Figure 1C) or edema/atrophy (P>0.05; Figure 1D) between the different experimental groups at 7 and 14 dpi. These results suggest that the functional recovery associated with the local and delayed delivery of PACAP must be correlated with modulation of late pathophysiological processes.
Delayed Local Delivery of PACAP Dampens Inflammatory Responses
To determine whether the local delivery of PACAP modulates the postischemic inflammation, we performed a transcriptomic analysis on an Openarray® platform using a panel of 632 genes involved in inflammation. This analysis was conducted at 7 dpi when functional recovery induced by ES-P–grafted cells was already effective. The differential transcriptomic profiles between ES-P and ES cell groups were then analyzed using IPA bioinformatic tools to determine regulated critical pathways that may participate in the neuroprotective action of PACAP. IPA analysis identified 5 main regulated gene networks, 4 of which having a significant score >30 (Figure 2). These 4 networks were related to the inflammatory response and encompassed genes involved in infectious disease, cellular movement, hematologic system development and function, immune cell trafficking, and cell death and survival (Figure 2A–2D). Interestingly, the results showed that compared with ES cell injections, ES-P cell grafts induced a modulation of chemotactic responses as illustrated by the downregulated expression of Cxcl9, Cxcl11, Ccl2, Ccrl1, Ccrl2, and Cxcr6, with a concomitant upregulated expression of Cxcl2, Cxcl6, Ccl5, Ccl3l1/Ccl3l3, Ccl8, Ccl25, Ccl13, Ccbp2, Ccr2, and Ccr7 (Figure 2A). In addition, IPA analyses highlighted a decrease of proinflammatory mediators as indicated by the downregulation of TNF (Figure 2D), interferons (Figure 2A), and IL-1 (Figure 2C) networks, in conjunction with an inhibition of the nuclear factor-κB pathway (Figure 2D). In parallel, the genes more likely involved in the resolution of the inflammatory process, such as Ptges, Pparg, Tgfb1, Hmox1, Twix1 (Figure 2B), and Tnfaip3 (Figure 2D), are upregulated.
PACAP-Dependent Immunomodulation Is Rapid and Stable
To confirm these results, we determined the expression levels of mRNAs encoding the proinflammatory cytokine TNF-α, the immunosuppressive cytokine IL-10, and the neuroprotective factor Ym-1 in ipsilateral hemispheres. These analyses were conducted at 7 and 14 dpi to study the evolution of the inflammatory response (Figure 3). Validating our panel-based transcriptomic results at 7 dpi, it was found that the expression level of TNF-α is significantly upregulated in the ischemic tissues of mice injected with saline or ES cells compared with that of sham-operated animals, whereas the TNF-α levels are significantly reduced in the brain of mice transplanted with ES-P cells (Figure 3A, left). Likewise, the expression levels of IL-10 and of Ym-1 are significantly increased in the ES-P group compared with other experimental groups (Figure 3B and 3C, left). Interestingly, at 14 dpi, the expression of TNF-α and IL-10 is maintained, respectively, at lower and higher levels in the ES-P cell group compared with the other groups (Figure 3A and 3B, right), whereas the expression level of Ym-1 returns back to sham levels (Figure 3C, right).
PACAP Regulation of Distinct Transcription Factors May Underlie Its Immunomodulatory Actions
From differential gene expression, IPA analysis identified several upstream regulators whose activation or inhibition may account for the observed PACAP-mediated responses. Thus, IPA identified 5 main activated transcription factors (Table), including CCAAT/enhancer-binding protein beta (CEBPB, z-score=2.129), early growth response protein 1 (EGR1, z-score=2.879), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (NFKBIA, z-score=2.233), signal transducer and activator of transcription 3 (STAT3, z-score=2.089), and transcription factor p65 (RELA, z-score=2.364), and 1 inhibited transcription factor recombination signal binding protein for immunoglobulin kappa J region (RBP-J, z-score=−2.417).
Functional Recovery Observed After Delayed PACAP Delivery Correlates With Polarization of Microglial Cells Toward a M2 Phenotype
Because transcriptomic and bioinformatic analyses demonstrate a correlation between functional recovery and modulation of the local inflammatory responses, we examined microglia/macrophage dynamics that are central to this process. We used Cx3cr1+/GFP heterozygous transgenic mice to explore specifically the activation status of microglial cells.18 Morphometric analysis of the GFP+ cells at 7 dpi in 2 peri-infarct regions (Figure 4A) revealed that the delayed PACAP delivery induced a significant decrease in the number of microglia/macrophage cells (Figure 4B, upper). Moreover, the area of the cell soma was reduced (Figure 4B, upper), whereas the number and area of fiber-like extension processes per cell (Figure 4B, lower left and right) were increased, suggesting that the phenotype of activated microglial/macrophage cells was altered. To confirm the PACAP-dependent skewing of the microglial response, we performed immunohistofluorescence labeling for the M2-phenotypic marker Arg-1 (Figure 4C and 4D). Because no GFP+ ES or ES-P cells seem Arg-1+ 1 week after transplantation in nontransgenic C57Bl/6 mice (Figure II in the online-only Data Supplement), we assume that the observed GFP+/Arg-1+ double positive cells derived only from host cells and not from the cells transplanted in the Cx3cr1+/GFP model. The quantification of Arg-1+/GFP+ cells revealed that 18.9±1.8% of GFP+ microglial/macrophage cells in the ES-P group are Arg-1+, whereas double-labeled cells represent only 6.1±0.6% and 6.3±0.8% of microglial/macrophage cells in the saline and ES cell–treated groups, respectively (*P<0.05; Figure 4D). The increased number of M2-phenotypic marker Arg-1 expressing microglial/macrophage cells in the border zone of the infarct area in mice transplanted with ES-P cells compared with the other groups (Merge channel; Figure 4C and 4D) further confirmed skewing of the microglial response.
Cerebral ischemia initiates a complex set of pathophysiological events that evolve over time and space, leading to a massive and progressive neurodegeneration promoting severe functional deficits. Except for thrombolysis, therapeutic strategies targeting the acute phase of stroke are still ineffective in patients. Thus, it becomes essential to develop new approaches targeting new therapeutic windows to significantly improve the processes of functional recovery.
The beneficial effects of PACAP infusion and stem cell transplantation are well documented for stroke treatment.11,19 Nevertheless, there are several hurdles that need to be addressed before these approaches can be used in therapy. To circumvent these limitations, we integrated the 2 approaches by establishing a stem cell line designed to produce and release PACAP (Figure I in the online-only Data Supplement). From the capacity of transplanted stem cells to migrate toward ischemic area, the local PACAP delivery skirts issues related to PACAP stability and targeting to damaged brain regions where vascularization has been compromised. Because PACAP also has differentiation promoting properties, its paracrine and autocrine actions may reduce stem cell tumorigenic potential and associated risks (Figure I in the online-only Data Supplement).
Numerous studies report beneficial effects of stem cell transplantation after stroke because of their ability to produce antiapoptotic, neurotrophic, and immunomodulatory factors.20 In our model, the transplantation of nonmodified stem cells does not promote functional recovery at 7 and 14 dpi, suggesting that these cells and their secreted products, transplanted at 3 dpi, are ineffective in reducing the functional deficits. On the contrary, delayed injection of PACAP-expressing stem cells promotes fast, stable, and efficient functional recovery, indicating that the neurological improvements result specifically from PACAP. In this experimental model of brain ischemia, the maximal infarct volume is reached 24 hours after middle cerebral artery occlusion and remains unchanged over time. As expected, the observed functional recovery did not correlate with a reduction of the lesion size or accelerated resorption of the edema. Moreover, the beneficial effects of ES-P cell transplantation do not rely on cell replacement by grafted cells. Indeed, many grafted cells easily visualized in the vicinity of the infarct zone at 7 dpi (Figure III in the online-only Data Supplement) were no longer detectable by day 14. Even in the absence of graft rejection reaction and local PACAP release, the survival of stem cells transplanted at 3 dpi seems strongly compromised, suggesting that the neuroprotective effect of the delayed PACAP delivery is not linked to its antiapoptotic properties. Despite the rapid disappearance of ES-P cells, the fast and stable neurological improvements observed after local but transient PACAP delivery suggest that PACAP is able to quickly and stably restore a cerebral environment compatible with neuronal survival and functions.
Among the pathophysiological events induced by brain ischemia, inflammatory processes contribute to neurological deficits by generating a neurotoxic environment and altering the activity of still viable neuronal networks. The present experimental model is based on stem cell transplantation in syngeneic animals to obviate immunosuppressive treatment and allow a reliable evaluation of the immunomodulatory properties of PACAP. Our results show that on delayed administration, stem cell–mediated PACAP delivery can still modulate the ongoing inflammatory response. As reported in previous work, the local delivery of PACAP decreases the expression of proinflammatory mediators, such as TNF-α and IL-1β, and increases the expression of factors related to the resolution of inflammation, such as IL-10, transforming growth factor-β, IL1R antagonist, or Ym-1, clearly establishing PACAP as a potent immunomodulator.8,10,21 Bioinformatic analysis of gene regulations from a panel of 632 genes involved in inflammatory processes has revealed networks of regulated genes that support the anti-inflammatory activity of the peptide. Noticeably, immune cell trafficking and hematological system development and function are significantly affected processes. PACAP has already been shown to modulate the profile of chemokines produced by activated immune cells and consequently to modulate differentially the recruitment of various subtypes of lymphocytes.22–25 The coordinate regulation of expression of Cxcl9, Cxcl11, and Ccl22, for example, could result in reduced Th1 and Th2 cell infiltration. Relatedly, the upregulation of Ccr7 expression and the local increase of transforming growth factor-β and IL-10 expression suggest the preferential recruitment of Treg cells that could account for some of the observed neuroprotective effects of PACAP. Similarly, based on the differential transcriptional signatures between ES-P and ES cell–injected groups, IPA transcriptomic analyses reveal an increase of angiogenesis-related processes, specifically in PACAP-treated animals. This PACAP effect is correlated with the increase of vascular endothelial growth factor C expression (fold change=66.7) in line with previous work, showing indirect PACAP-mediated control of endothelial cell proliferation through the induction of vascular endothelial growth factor production.26,27 Because CD11b+ cells represent the major source of vascular endothelial growth factor C in the brain parenchyma,28 our data indicate that the beneficial effects of delayed PACAP delivery could rely on PACAP-induced vascular endothelial growth factor C–producing microglia/macrophages, thus promoting postischemic neovascularization. This hypothesis is reinforced by the increased Arg-1 labeling in vessel-like structures, illustrating an increased neoangiogenesis in ES-P cell–transplanted mice specifically. The fact that the local delivery of PACAP at 3 dpi could target the microglia/macrophage compartment, which mainly supports the early phase of the postischemic inflammatory response, is also substantiated by our transcriptomic analysis reporting that local PACAP release increases phagocytic activity and survival of microglia/macrophages (Table I in the online-only Data Supplement). This was further confirmed by our immunohistofluorescence experiments, demonstrating morphological changes of ischemia-activated microglia/macrophages and increased number of cells expressing the M2-phenotypic marker Arg-1 in the infarct border zone, specifically in ES-P–grafted mice. Noticeably, the close proximity of Arg-1+ vessels and Arg-1+/GFP+ microglia/macrophage cells may suggest a local and direct involvement of M2-like microglia/macrophages in the neoangiogenic process. Altogether, these results tend to demonstrate that the PACAP-dependent functional recovery improvements rely on the redirection of the microglial response toward a M2-like phenotype. After stroke and in parallel of microglia activation, many inflammatory cells infiltrate the brain parenchyma. Initially, monocyte-derived macrophages and resident microglial cells were both considered as detrimental. However, more recently, studies demonstrated that these 2 cell populations are functionally distinct, engaged in nonredundant roles.29 More particularly, Gliem et al30 showed that the inflammatory Ly6ChiCX3CR1intCCR2+ monocytes are protective after stroke by preventing hemorrhagic transformation and delaying clinical deterioration.30 This suggests that the reduction of the functional deficits observed after delayed PACAP delivery may result from an increased recruitment of inflammatory monocytes, exerting an activity of inflammation resolution. Nevertheless, because injection of M2-differentiated macrophages 4 days after stroke does not improve functional recovery in contrast to the transplantation of human microglial cells,31,32 and because delayed PACAP delivery modulates the tissular chemotactic response, we think that the improved functional recovery results more probably from the redirection of the microglial response toward a neuroprotective Mi2 phenotype.
Interestingly, in a murine model of permanent middle cerebral occlusion, Perego et al33 reported, 24 hours after ischemia, that the microglia/macrophages adopt a M2-like phenotype that progressively evolves toward a neurotoxic phenotype at day 7 post stroke. Because the kinetics of expression of endogenous PACAP after stroke by pyramidal cortical neurons (PACAP expression increases as soon as 6 hours after middle cerebral artery occlusion and peaks at 24 hours before disappearing) parallels with the evolution of the microglial response,34 the in vivo PACAP release can be part of an endogenous protective mechanism aimed at controlling the differentiation process of microglial cells, the loss of PACAP favoring the acquisition of a neurotoxic phenotype.
Finally, our study identifies the NFKBIA (IκB-α), C/EBP-β, and RBP-J transcriptional regulators as factors potentially mediating PACAP’s polarizing effects on microglia/macrophages (Figure IV in the online-only Data Supplement). The inhibition of nuclear factor-κB pathway has been already reported to mediate the anti-inflammatory properties of the neuropeptide.14,35 PACAP-dependent inhibition of IKK-β phosphorylation induces NFKBIA (IκB-α) stabilization, thereby decreasing nuclear factor-κB-p65 nuclear translocation. In macrophages, the neuropeptide PACAP induces through vasoactive intestinal peptide receptor 1 activation, cyclic adenosine monophosphate response element–binding protein (CREB) phosphorylation which in turn induces the expression of C/EBP-β.35,36 Interestingly, CREB-C/EBP-β cascade has been reported to support the expression of M2 macrophage–specific gene expression.36 In association with the decrease of RBP-J activity, a factor recently associated with M1 differentiation processes in macrophages37; these results suggest that PACAP could induce a shift toward the M2 microglial phenotype through CREB and RBP-J regulation.
In conclusion, we show that even delayed PACAP local delivery can efficiently promote functional recovery after brain stroke. The PACAP-dependent neurological improvements are associated with the modulation of inflammation and, more precisely, of the microglia/macrophage responses. The ability to redirect the microglia/macrophage phenotype from a M1 phenotype toward a M2 response by PACAP delivery at 3 dpi highlights the enduring plasticity of microglial cells after the stroke onset. Further studies are required to evaluate the PACAP-dependent modulation of the Notch/RBP-J pathway in microglial cells and to decipher the mechanisms by which M2-like microglial cells can improve functional recovery after stroke. It seems important to evaluate the functional relationships among the different microglia populations and their roles in cerebral and neural plasticity. Nevertheless, our results confirm the importance of developing approaches designed to target the local inflammatory response to open new therapeutical windows for stroke treatment. In this context, immunomodulatory strategies capable of redirecting the microglial response toward the neuroprotective M2 phenotype could represent attractive options for stroke treatment and perhaps even more widely for neurodegenerative disease with inflammatory components.
We thank Tommy Seaborn for critically reading the article and Pascal Hilber for advices in behavioral studies.
Sources of Funding
Dr Brifault was a recipient of a doctoral fellowship from the Région Haute-Normandie. This study was supported by Institut National de la Santé et de la Recherche Médicale (U982), Rouen University, Plate-Forme de Recherche en Imagerie Cellulaire de Haute-Normandie, and the Interreg Peptide Research Network of Excellence project.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.006864/-/DC1.
- Received July 22, 2014.
- Revision received November 10, 2014.
- Accepted November 28, 2014.
- © 2014 American Heart Association, Inc.
- Lansberg MG,
- Bluhmki E,
- Thijs VN
- Reglodi D,
- Somogyvari-Vigh A,
- Vigh S,
- Kozicz T,
- Arimura A
- Delgado M,
- Ganea D
- Delgado M,
- Munoz-Elias EJ,
- Gomariz RP,
- Ganea D
- Ohtaki H,
- Nakamachi T,
- Dohi K,
- Aizawa Y,
- Takaki A,
- Hodoyama K,
- et al
- Tsenter J,
- Beni-Adani L,
- Assaf Y,
- Alexandrovich AG,
- Trembovler V,
- Shohami E
- Jung S,
- Aliberti J,
- Graemmel P,
- Sunshine MJ,
- Kreutzberg GW,
- Sher A,
- et al
- Bacigaluppi M,
- Pluchino S,
- Peruzzotti-Jametti L,
- Kilic E,
- Kilic U,
- Salani G,
- et al
- Delgado M,
- Gonzalez-Rey E,
- Ganea D
- Gloddek J,
- Pagotto U,
- Paez Pereda M,
- Arzt E,
- Stalla GK,
- Renner U
- Lalancette-Hébert M,
- Gowing G,
- Simard A,
- Weng YC,
- Kriz J
- Stumm R,
- Kolodziej A,
- Prinz V,
- Endres M,
- Wu DF,
- Höllt V
- Delgado M,
- Ganea D
- Ruffell D,
- Mourkioti F,
- Gambardella A,
- Kirstetter P,
- Lopez RG,
- Rosenthal N,
- et al