Embodied bidirectional simulation of a spiking cortico-basal ganglia-cerebellar-thalamic brain model and a mouse musculoskeletal body model distributed across computers including the supercomputer Fugaku.

Embodied simulation with a digital brain model and a realistic musculoskeletal body model provides a means to understand animal behavior and behavioral change. Such simulation can be too large and complex to conduct on a single computer, and so distributed simulation across multiple computers over the Internet is necessary. In this study, we report our joint effort on developing a spiking brain model and a mouse body model, connecting over the Internet, and conducting bidirectional simulation while synchronizing them. Specifically, the brain model consisted of multiple regions including secondary motor cortex, primary motor and somatosensory cortices, basal ganglia, cerebellum and thalamus, whereas the mouse body model, provided by the Neurorobotics Platform of the Human Brain Project, had a movable forelimb with three joints and six antagonistic muscles to act in a virtual environment. Those were simulated in a distributed manner across multiple computers including the supercomputer Fugaku, which is the flagship supercomputer in Japan, while communicating via Robot Operating System (ROS). To incorporate models written in C/C++ in the distributed simulation, we developed a C++ version of the rosbridge library from scratch, which has been released under an open source license. These results provide necessary tools for distributed embodied simulation, and demonstrate its possibility and usefulness toward understanding animal behavior and behavioral change.

Jejunal gastrointestinal stromal tumor that developed in a patient with neurofibromatosis type 1: a case report.

BACKGROUND: Neurofibromatosis type 1 (NF1) is known to be associated with the frequent occurrence of unique gastrointestinal stromal tumors (GISTs), preferably occurring in the small intestine, with no mutations in the c-kit proto-oncogene or platelet-derived growth factor receptor-alpha (PDGFRA), with a high tendency for multifocal development, indolent nature, with low proliferation activity and favorable prognosis. CASE PRESENTATION: A woman in her forties visited her local doctor complaining of menstrual pain; a large mass was detected in her lower abdomen, and she was referred to our hospital. The patient had hundreds of skin warts and café au lait spots. The patient's mother had been diagnosed with type 1 neurofibromatosis. The patient met the diagnostic criteria for NF1 and was diagnosed with NF1. Ultrasonography showed a large heterogeneous cystic mass with various echo patterns, solid compartments and multiple septations. Magnetic resonance imaging showed a multilocular cystic mass with liquid content exhibiting various intensities, including that of blood. A small round solid mass was also observed close to the cystic tumor. Contrast-enhanced computed tomography showed that the round solid mass showed strong enhancement in the early phase, unlike the cystic tumor component. Open laparotomy revealed a multicystic exophytic tumor measuring 11.5 cm originating from the jejunal wall, 20 cm distal to the duodenojejunal flexure. A solid tumor measuring 2.1 cm was also found on the anal side of the large tumor. We resected the short segment of the jejunum, including the two lesions. Microscopic findings revealed that the cystic and solid tumors consisted of spindle-shaped tumor cells showing little atypia with a fascicular or bundle arrangement. Nuclear mitosis was scarce. Immunostaining of the tumor cells showed positive staining for KIT and DOG1 and negative staining for S100 and desmin. The NF1 patient was diagnosed with multiple GISTs accompanied by intratumoral hemorrhagic denaturation arising from the jejunum. The TNM staging was pT4N0M0, stage IIIA. CONCLUSION: We report a case of GISTs associated with NF1 that showed a jejunal origin, multifocal development and few mitotic figures. The recurrence risk, survival prognosis and need for adjuvant chemotherapy, particularly in cases where the initial GIST exhibits a very indolent pathology in NF1-related GISTs, remain to be elucidated.

Histological transformation to signet-ring cell carcinoma in a patient with clinically aggressive poorly differentiated adenocarcinoma of the ascending colon after response to chemotherapy plus cetuximab: a case report.

BACKGROUND: Alteration of chemosensitivity or tumor aggressiveness in response to chemotherapy has been reported, and liquid biopsy assessment during chemotherapy for colorectal cancers has confirmed the acquisition of mutations in various oncogenes. However, the occurrence of histological transformation seems to be extremely rare in colorectal cancers, and the few existing case reports of this transformation are from lung cancer and breast cancer. In this report, we describe the histological transformation of clinically aggressive scirrhous-type poorly differentiated adenocarcinoma of the ascending colon to signet-ring cell carcinoma in almost all recurrent tumors that were confirmed by autopsy after response to chemotherapy plus cetuximab. CASE PRESENTATION: A 59-year-old woman visited our hospital with whole abdominal pain and body weight loss and was diagnosed with scirrhous-type poorly differentiated adenocarcinoma of the ascending colon with aggressive lymph node metastases. The intrinsic chemosensitivity of the tumors was evident upon initiation of mFOLFOX6 plus cetuximab therapy, and right hemicolectomy was performed, and the tumor obviously remained in the peripancreatic area, paraaortic region, or other retroperitoneal areas. The ascending colon tumors mainly consisted of poorly differentiated adenocarcinoma and were not associated with signet-ring cell components except for minute clusters in a few lymphatic emboli in the main tumor. Chemotherapy was continued, and metastases were eliminated at 8 months after the operation; this response was maintained for an additional 4 months. Discontinuation of chemotherapy plus cetuximab resulted in immediate tumor recurrence and rapid expansion, and the patient died of the recurrent tumor 1 year and 2 months after the operation. Autopsy specimens revealed that almost all of the recurrent tumors exhibited transformation and consisted of signet-ring cell histology. CONCLUSION: This case might suggest that various oncogene mutations or epigenetic changes resulting from chemotherapy, especially regimens that include cetuximab, contribute to the transformation of non-signet-ring cell colorectal carcinoma to signet-ring cell carcinoma histology and can promote the aggressive clinical progression characteristic of signet-ring cell carcinoma.

Deploying and Optimizing Embodied Simulations of Large-Scale Spiking Neural Networks on HPC Infrastructure.

Simulating the brain-body-environment trinity in closed loop is an attractive proposal to investigate how perception, motor activity and interactions with the environment shape brain activity, and vice versa. The relevance of this embodied approach, however, hinges entirely on the modeled complexity of the various simulated phenomena. In this article, we introduce a software framework that is capable of simulating large-scale, biologically realistic networks of spiking neurons embodied in a biomechanically accurate musculoskeletal system that interacts with a physically realistic virtual environment. We deploy this framework on the high performance computing resources of the EBRAINS research infrastructure and we investigate the scaling performance by distributing computation across an increasing number of interconnected compute nodes. Our architecture is based on requested compute nodes as well as persistent virtual machines; this provides a high-performance simulation environment that is accessible to multi-domain users without expert knowledge, with a view to enable users to instantiate and control simulations at custom scale via a web-based graphical user interface. Our simulation environment, entirely open source, is based on the Neurorobotics Platform developed in the context of the Human Brain Project, and the NEST simulator. We characterize the capabilities of our parallelized architecture for large-scale embodied brain simulations through two benchmark experiments, by investigating the effects of scaling compute resources on performance defined in terms of experiment runtime, brain instantiation and simulation time. The first benchmark is based on a large-scale balanced network, while the second one is a multi-region embodied brain simulation consisting of more than a million neurons and a billion synapses. Both benchmarks clearly show how scaling compute resources improves the aforementioned performance metrics in a near-linear fashion. The second benchmark in particular is indicative of both the potential and limitations of a highly distributed simulation in terms of a trade-off between computation speed and resource cost. Our simulation architecture is being prepared to be accessible for everyone as an EBRAINS service, thereby offering a community-wide tool with a unique workflow that should provide momentum to the investigation of closed-loop embodiment within the computational neuroscience community.

Human-scale Brain Simulation via Supercomputer: A Case Study on the Cerebellum.

Performance of supercomputers has been steadily and exponentially increasing for the past 20 years, and is expected to increase further. This unprecedented computational power enables us to build and simulate large-scale neural network models composed of tens of billions of neurons and tens of trillions of synapses with detailed anatomical connections and realistic physiological parameters. Such "human-scale" brain simulation could be considered a milestone in computational neuroscience and even in general neuroscience. Towards this milestone, it is mandatory to introduce modern high-performance computing technology into neuroscience research. In this article, we provide an introductory landscape about large-scale brain simulation on supercomputers from the viewpoints of computational neuroscience and modern high-performance computing technology for specialists in experimental as well as computational neurosciences. This introduction to modeling and simulation methods is followed by a review of various representative large-scale simulation studies conducted to date. Then, we direct our attention to the cerebellum, with a review of more simulation studies specific to that region. Furthermore, we present recent simulation results of a human-scale cerebellar network model composed of 86 billion neurons on the Japanese flagship supercomputer K (now retired). Finally, we discuss the necessity and importance of human-scale brain simulation, and suggest future directions of such large-scale brain simulation research.

Simulation of a Human-Scale Cerebellar Network Model on the K Computer.

Computer simulation of the human brain at an individual neuron resolution is an ultimate goal of computational neuroscience. The Japanese flagship supercomputer, K, provides unprecedented computational capability toward this goal. The cerebellum contains 80% of the neurons in the whole brain. Therefore, computer simulation of the human-scale cerebellum will be a challenge for modern supercomputers. In this study, we built a human-scale spiking network model of the cerebellum, composed of 68 billion spiking neurons, on the K computer. As a benchmark, we performed a computer simulation of a cerebellum-dependent eye movement task known as the optokinetic response. We succeeded in reproducing plausible neuronal activity patterns that are observed experimentally in animals. The model was built on dedicated neural network simulation software called MONET (Millefeuille-like Organization NEural neTwork), which calculates layered sheet types of neural networks with parallelization by tile partitioning. To examine the scalability of the MONET simulator, we repeatedly performed simulations while changing the number of compute nodes from 1,024 to 82,944 and measured the computational time. We observed a good weak-scaling property for our cerebellar network model. Using all 82,944 nodes, we succeeded in simulating a human-scale cerebellum for the first time, although the simulation was 578 times slower than the wall clock time. These results suggest that the K computer is already capable of creating a simulation of a human-scale cerebellar model with the aid of the MONET simulator.

Large-Scale Simulation of a Layered Cortical Sheet of Spiking Network Model Using a Tile Partitioning Method.

One of the grand challenges for computational neuroscience and high-performance computing is computer simulation of a human-scale whole brain model with spiking neurons and synaptic plasticity using supercomputers. To achieve such a simulation, the target network model must be partitioned onto a number of computational nodes, and the sub-network models are executed in parallel while communicating spike information across different nodes. However, it remains unclear how the target network model should be partitioned for efficient computing on next generation of supercomputers. Specifically, reducing the communication of spike information across compute nodes is essential, because of the relatively slower network performance than processor and memory. From the viewpoint of biological features, the cerebral cortex and cerebellum contain 99% of neurons and synapses and form layered sheet structures. Therefore, an efficient method to split the network should exploit the layered sheet structures. In this study, we indicate that a tile partitioning method leads to efficient communication. To demonstrate it, a simulation software called MONET (Millefeuille-like Organization NEural neTwork simulator) that partitions a network model as described above was developed. The MONET simulator was implemented on the Japanese flagship supercomputer K, which is composed of 82,944 computational nodes. We examined a performance of calculation, communication and memory consumption in the tile partitioning method for a cortical model with realistic anatomical and physiological parameters. The result showed that the tile partitioning method drastically reduced communication data amount by replacing network communication with DRAM access and sharing the communication data with neighboring neurons. We confirmed the scalability and efficiency of the tile partitioning method on up to 63,504 compute nodes of the K computer for the cortical model. In the companion paper by Yamaura et al., the performance for a cerebellar model was examined. These results suggest that the tile partitioning method will have advantage for a human-scale whole-brain simulation on exascale computers.

Corrigendum: Extremely Scalable Spiking Neuronal Network Simulation Code: From Laptops to Exascale Computers.

[This corrects the article DOI: 10.3389/fninf.2018.00002.].

First case of a bloodstream infection caused by the genus Brachybacterium.

An 83-year-old previously self-sufficient man was referred to our hospital for a fever, severe tenderness over the lumbar spine, and elevated C-reactive protein levels. Computed tomography revealed fluid collection in the intervertebral space of L3/4. Gram-positive, short rod-shaped bacteria were isolated from two sets of blood cultures. A 16S rRNA sequence analysis of an isolate showed a similarity of 98.1% to the nearest type strain Brachybacterium squillarum JCM 16464T. Biochemical characteristics of the presently isolated strain differed from those of the most closely related species of the genus Brachybacterium. The patient was successfully discharged on day 73 of admission with antimicrobial therapies and showed no recurrence during outpatient visits. Brachybacterium spp. have mainly been isolated from the environment, and human Brachybacterium infections have rarely been documented to date. To our knowledge, this is the first clinical isolation of Brachybacterium sp. as a causative pathogen of bloodstream infection.

Extremely Scalable Spiking Neuronal Network Simulation Code: From Laptops to Exascale Computers.

State-of-the-art software tools for neuronal network simulations scale to the largest computing systems available today and enable investigations of large-scale networks of up to 10 % of the human cortex at a resolution of individual neurons and synapses. Due to an upper limit on the number of incoming connections of a single neuron, network connectivity becomes extremely sparse at this scale. To manage computational costs, simulation software ultimately targeting the brain scale needs to fully exploit this sparsity. Here we present a two-tier connection infrastructure and a framework for directed communication among compute nodes accounting for the sparsity of brain-scale networks. We demonstrate the feasibility of this approach by implementing the technology in the NEST simulation code and we investigate its performance in different scaling scenarios of typical network simulations. Our results show that the new data structures and communication scheme prepare the simulation kernel for post-petascale high-performance computing facilities without sacrificing performance in smaller systems.

Effects of an autoinducer analogue on antibiotic tolerance in Pseudomonas aeruginosa.

Objectives: Antibiotic tolerance causes chronic, refractory and persistent infections. In order to advance the development of a new type of drug for the treatment of infectious diseases, we herein investigated the effects of a newly synthesized analogue of the Pseudomonas aeruginosa quorum-sensing autoinducer named AIA-1 ( a uto i nducer a nalogue) on antibiotic tolerance in P. aeruginosa . Methods: A P. aeruginosa luminescent strain derived from PAO1 was injected into neutropenic ICR mice and bioluminescence images were acquired for a period of time after treatments with antibiotics and AIA-1. In vitro susceptibility testing and killing assays for the planktonic and biofilm cells of PAO1 were performed using antibiotics and AIA-1. The expression of quorum-sensing-related genes was examined using real-time PCR. Results: In vivo and in vitro assays showed that AIA-1 alone did not exert any bactericidal effects and also did not affect the MICs of antibiotics. However, the combined use of AIA-1 and antibiotics exerted markedly stronger therapeutic effects against experimental infection than antibiotics alone. The presence of AIA-1 also enhanced the killing effects of antibiotics in planktonic and biofilm cells. Although AIA-1 did not inhibit the expression of lasB and rhlA genes, which are directly regulated by quorum sensing, it clearly suppressed expression of the rpoS gene. Conclusions: The new compound, AIA-1, did not alter the antibiotic susceptibility of P. aeruginosa by itself; however, its addition enhanced the antibacterial activity of antibiotics. AIA-1 did not inhibit quorum sensing, but reduced the antibiotic tolerance of P. aeruginosa by suppressing rpoS gene expression.

A Pseudomonas aeruginosa Quorum-Sensing autoinducer analog enhances the activity of antibiotics against resistant strains.

In this study, we have investigated the effects of the newly synthesized analog of Pseudomonas aeruginosa quorum-sensing autoinducer named AIA-1 (autoinducer analog) against antibiotic-resistant bacteria. In vitro susceptibility and killing assays for P. aeruginosa PAO1ΔoprD mutant and clinical isolates were performed by using antibiotics and AIA-1. In an in vivo assay, a luminescent carbapenem-resistant strain derived from PAO1ΔoprD was injected into neutropenic ICR mice and bioluminescence images were acquired after the treatment with antibiotics and AIA-1. Additionally, we investigated the effects of the combination use against carbapenem-resistant Enterobacteriaceae (CRE). Using killing assays in P. aeruginosa, the survival rates in the presence of antibiotics and AIA-1 significantly decreased in comparison with those with antibiotics alone. Furthermore, dual treatment of biapenem and AIA-1 was more effective than biapenem alone in a mouse infection model. AIA-1 did not change the MICs in P. aeruginosa, suggesting that AIA-1 acts on the mechanism of antibiotic tolerance. Conversely, the MICs of antibiotics decreased in the presence of AIA-1 in some CRE strains, indicating that AIA-1 may require additional mechanism to act on CRE. In conclusion, AIA-1 may be a potent drug for clinical treatment of infections caused by antibiotic-resistant bacteria. J. Med. Invest. 64: 101-109, February, 2017.

Collective excitations in Na2IrO3.

We study the collective excitations of Na2IrO3 in an itinerant electron approach. We consider a multi-orbital tight-binding model with the electron transfer between the Ir 5d states mediated via oxygen 2p states and the direct d-d transfer on a honeycomb lattice. The one electron energy as well as the ground state energy are investigated within the Hartree-Fock approximation. When the direct d-d transfer is weak, we obtain nearly flat energy bands due to the formation of quasimolecular orbitals, and the ground state exhibits the zigzag spin order. The evaluation of the density-density correlation function within the random phase approximation shows that the collective excitations emerge as bound states. For an appropriate value of the direct d-d transfer, some of them are concentrated in the energy region ω<50 meV(magnetic excitations) while the others lie in the energy region ω>350 meV (excitonic excitations). This behaviour is consistent with the resonant inelastic x-ray scattering spectra. We also show that the larger values of the direct d-d transfer are unfavourable in order to explain the observed aspects of Na2IrO3 such as the ordering pattern of the ground state and the excitation spectrum. These findings may indicate that the direct d-d transfer is suppressed by the structural distortions in the view of excitation spectroscopy, as having been pointed out in the ab initio calculation.

A unified framework for spiking and gap-junction interactions in distributed neuronal network simulations.

Contemporary simulators for networks of point and few-compartment model neurons come with a plethora of ready-to-use neuron and synapse models and support complex network topologies. Recent technological advancements have broadened the spectrum of application further to the efficient simulation of brain-scale networks on supercomputers. In distributed network simulations the amount of spike data that accrues per millisecond and process is typically low, such that a common optimization strategy is to communicate spikes at relatively long intervals, where the upper limit is given by the shortest synaptic transmission delay in the network. This approach is well-suited for simulations that employ only chemical synapses but it has so far impeded the incorporation of gap-junction models, which require instantaneous neuronal interactions. Here, we present a numerical algorithm based on a waveform-relaxation technique which allows for network simulations with gap junctions in a way that is compatible with the delayed communication strategy. Using a reference implementation in the NEST simulator, we demonstrate that the algorithm and the required data structures can be smoothly integrated with existing code such that they complement the infrastructure for spiking connections. To show that the unified framework for gap-junction and spiking interactions achieves high performance and delivers high accuracy in the presence of gap junctions, we present benchmarks for workstations, clusters, and supercomputers. Finally, we discuss limitations of the novel technology.

Pararenal Lymphatic Cyst Infection Caused by Helicobacter cinaedi.

A 43-year-old man was referred to our hospital for an acute-onset fever and left flank pain. He had been previously diagnosed with lymphangioma, and abdominal computed tomography showed pararenal cysts with fat stranding around the left kidney, of which infection was subsequently confirmed on magnetic resonance imaging. Gram-negative spiral bacilli were isolated from two sets of blood cultures, and Helicobacter cinaedi was identified using 16S rRNA sequencing. The patient was successfully treated with ceftriaxone therapy without recurrence. A multilocus sequence typing analysis indicated the current H. cinaedi strain differed from previous strains isolated in Japan.

Preclinical Study of Novel Gene Silencer Pyrrole-Imidazole Polyamide Targeting Human TGF-ß1 Promoter for Hypertrophic Scars in a Common Marmoset Primate Model.

We report a preclinical study of a pyrrole-imidazole (PI) polyamide that targets the human transforming growth factor (hTGF)-ß1 gene as a novel transcriptional gene silencer in a common marmoset primate model. We designed and then synthesized PI polyamides to target the hTGF-ß1 promoter. We examined effects of seven PI polyamides (GB1101-1107) on the expression of hTGF-ß1 mRNA stimulated with phorbol 12-myristate 13-acetate (PMA) in human vascular smooth muscle cells. GB1101, GB1105 and GB1106 significantly inhibited hTGF-ß1 mRNA expression. We examined GB1101 as a PI polyamide to hTGF-ß1 for hypertrophic scars in marmosets in vivo. Injection of GB1101 completely inhibited hypertrophic scar formation at 35 days post-incision and inhibited cellular infiltration, TGF-ß1 and vimentin staining, and epidermal thickness. Mismatch polyamide did not affect hypertrophic scarring or histological changes. Epidermis was significantly thinner with GB1101 than with water and mismatch PI polyamides. We developed the PI polyamides for practical ointment medicines for the treatment of hypertrophic scars. FITC-labeled GB1101 with solbase most efficiently distributed in the nuclei of epidermal keratinocytes, completely suppressed hypertropic scarring at 42 days after incision, and considerably inhibited epidermal thickness and vimentin-positive fibroblasts. PI polyamides targeting hTGF-ß1 promoter with solbase ointment will be practical medicines for treating hypertrophic scars after surgical operations and skin burns.

Effect of broken symmetry on resonant inelastic x-ray scattering from undoped cuprates.

We study the magnetic excitation spectra of resonant inelastic x-ray scattering (RIXS) at the L-edge from undoped cuprates beyond the fast collision approximation. We analyse the effect of the symmetry breaking ground state on the RIXS process of the Heisenberg model by using a projection procedure. We derive the expressions of the scattering amplitude in both one-magnon and two-magnon excitation channels. Each of them consists of the isotropic and anisotropic contributions. The latter is a new finding and attributed to the long range order of the ground state. The presence of anisotropic terms is supported by numerical calculations on a two-dimensional spin cluster. We express the RIXS spectra in the form of spin-correlation functions with the coefficients evaluated on the cluster, and calculate the function in a two dimensional system within the 1/S expansion. Due to the anisotropic terms, the spectral intensities are considerably enhanced around momentum transfer q = 0 in both one-magnon and two-magnon excitation channels. This finding may be experimentally confirmed by examining carefully the q-dependence of the spectra.

Identification of a novel E-box binding pyrrole-imidazole polyamide inhibiting MYC-driven cell proliferation.

The MYC transcription factor plays a crucial role in the regulation of cell cycle progression, apoptosis, angiogenesis, and cellular transformation. Due to its oncogenic activities and overexpression in a majority of human cancers, it is an interesting target for novel drug therapies. MYC binding to the E-box (5'-CACGTGT-3') sequence at gene promoters contributes to more than 4000 MYC-dependent transcripts. Owing to its importance in MYC regulation, we designed a novel sequence-specific DNA-binding pyrrole-imidazole (PI) polyamide, Myc-5, that recognizes the E-box consensus sequence. Bioinformatics analysis revealed that the Myc-5 binding sequence appeared in 5'- MYC binding E-box sequences at the eIF4G1, CCND1, and CDK4 gene promoters. Furthermore, ChIP coupled with detection by quantitative PCR indicated that Myc-5 has the ability to inhibit MYC binding at the target gene promoters and thus cause downregulation at the mRNA level and protein expression of its target genes in human Burkitt's lymphoma model cell line, P493.6, carrying an inducible MYC repression system and the K562 (human chronic myelogenous leukemia) cell line. Single i.v. injection of Myc-5 at 7.5 mg/kg dose caused significant tumor growth inhibition in a MYC-dependent tumor xenograft model without evidence of toxicity. We report here a compelling rationale for the identification of a PI polyamide that inhibits a part of E-box-mediated MYC downstream gene expression and is a model for showing that phenotype-associated MYC downstream gene targets consequently inhibit MYC-dependent tumor growth.

Identification of frequent differentially methylated region in sporadic bladder cancers.

INTRODUCTION: Aberrant methylation levels in the cytosine-phosphate-guanine island (CpGi) region from exon 1 to intron 1 of the zygote arrest 1 (ZAR1) gene have been reported in several types of human cancers, including melanoma, brain tumor, and hepatocellular carcinoma. In the present study, methylation levels at the CpGi of ZAR1 exon 1/intron 1 in bladder cancer specimens were analyzed using mass spectrometry. MATERIALS AND METHODS: Genomic DNA was extracted from 20 sporadic bladder cancers, and the methylation levels at ZAR1 CpGi were quantitatively examined by the MassARRAY EpiTYPER method. RESULT: The methylation levels at specific CpG sites of the ZAR1 CpGi were significantly lower in high-grade bladder cancers than in low-grade tumors. CONCLUSIONS: The results of the present study indicated a decreased methylation level at CpG sites of ZAR1 exon 1/intron 1. CpGi could serve as a biomarker for invasive bladder cancer.

Spiking network simulation code for petascale computers.

Brain-scale networks exhibit a breathtaking heterogeneity in the dynamical properties and parameters of their constituents. At cellular resolution, the entities of theory are neurons and synapses and over the past decade researchers have learned to manage the heterogeneity of neurons and synapses with efficient data structures. Already early parallel simulation codes stored synapses in a distributed fashion such that a synapse solely consumes memory on the compute node harboring the target neuron. As petaflop computers with some 100,000 nodes become increasingly available for neuroscience, new challenges arise for neuronal network simulation software: Each neuron contacts on the order of 10,000 other neurons and thus has targets only on a fraction of all compute nodes; furthermore, for any given source neuron, at most a single synapse is typically created on any compute node. From the viewpoint of an individual compute node, the heterogeneity in the synaptic target lists thus collapses along two dimensions: the dimension of the types of synapses and the dimension of the number of synapses of a given type. Here we present a data structure taking advantage of this double collapse using metaprogramming techniques. After introducing the relevant scaling scenario for brain-scale simulations, we quantitatively discuss the performance on two supercomputers. We show that the novel architecture scales to the largest petascale supercomputers available today.