Computational models are complex scientific constructs that have become essential for us to better understand the world. Many models are valuable for peers within and beyond disciplinary boundaries. However, there are no widely agreed-upon standards for sharing models. This paper suggests 10 simple rules for you to both (i) ensure you share models in a way that is at least "good enough," and (ii) enable others to lead the change towards better model-sharing practices.

Constitutively secreted by endothelial cells, soluble FLT1 (sFLT1 or sVEGFR1) binds and sequesters extracellular vascular endothelial growth factors (VEGF), thereby reducing VEGF binding to VEGF receptor tyrosine kinases and their downstream signaling. In doing so, sFLT1 plays an important role in vascular development and in the patterning of new blood vessels in angiogenesis. Here, we develop multiple mechanistic models of sFLT1 secretion and identify a minimal mechanistic model that recapitulates key qualitative and quantitative features of temporal experimental datasets of sFLT1 secretion from multiple studies. We show that the experimental data on sFLT1 secretion is best represented by a delay differential equation (DDE) system including a maturation term, reflecting the time required between synthesis and secretion. Using optimization to identify appropriate values for the key mechanistic parameters in the model, we show that two model parameters (extracellular degradation rate constant and maturation time) are very strongly constrained by the experimental data, and that the remaining parameters are related by two strongly constrained constants. Thus, only one degree of freedom remains, and measurements of the intracellular levels of sFLT1 would fix the remaining parameters. Comparison between simulation predictions and additional experimental data of the outcomes of chemical inhibitors and genetic perturbations suggest that intermediate values of the secretion rate constant best match the simulation with experiments, which would completely constrain the model. However, some of the inhibitors tested produce results that cannot be reproduced by the model simulations, suggesting that additional mechanisms not included here are required to explain those inhibitors. Overall, the model reproduces most available experimental data and suggests targets for further quantitative investigation of the sFLT1 system.

The vascular endothelial growth factor receptors (VEGFRs) bind to cognate ligands to facilitate signaling pathways critical for angiogenesis, the growth of new capillaries from existing vasculature. Intracellular trafficking regulates the availability of receptors on the cell surface to bind ligands, which regulate activation, and the movement of activated receptors between the surface and intracellular pools, where they can initiate different signaling pathways. Using experimental data and computational modeling, we recently demonstrated and quantified the differential trafficking of three VEGF receptors, VEGFR1, VEGFR2, and coreceptor Neuropilin-1 (NRP1). Here, we expand that approach to quantify how the binding of different VEGF ligands alters the trafficking of these VEGF receptors and demonstrate the consequences of receptor localization and ligand binding on the localization and dynamics of signal initiation complexes. We include simulations of four different splice isoforms of VEGF-A and PLGF, each of which binds to different combinations of the VEGF receptors, and we use new experimental data for two of these ligands to parameterize and validate our model. We show that VEGFR2 trafficking is altered in response to ligand binding, but that trafficking of VEGFR1 is not; we also show that the altered trafficking can be explained by a single mechanistic process, increased internalization of the VEGFR2 receptor when bound to ligand; other processes are unaffected. We further show that even though the canonical view of receptor tyrosine kinases is of activation on the cell surface, most of the ligand-receptor complexes for both VEGFR1 and VEGFR2 are intracellular. We also explore the competition between the receptors for ligand binding, the so-called 'decoy effect', and show that while in vitro on the cell surface minimal such effect would be observed, inside the cell the effect can be substantial and may influence signaling. We term this location dependence the 'reservoir effect' as the size of the local ligand reservoir (large outside the cell, small inside the cell) plays an integral role in the receptor-receptor competition. These results expand our understanding of receptor-ligand trafficking dynamics and are critical for the design of therapeutic agents to regulate ligand availability to VEGFR1 and hence VEGF receptor signaling in angiogenesis.

My first exposure to visibly fluorescent proteins (FPs) was near the end of my time as a faculty member at the University of California, Berkeley.Prof. Alexander Glazer, a friend and colleague there, was the world's expert on phycobiliproteins, the brilliantly colored and intensely fluorescent proteins that serve as light-harvesting antennae for the photosynthetic apparatus of blue-green algae or cyanobacteria.One day, probably around 1987-88, Glazer told me that his lab had cloned the gene for one of the phycobiliproteins.Furthermore, he said, the apoprotein produced from this gene became fluorescent when mixed with its chromophore, a small molecule cofactor that could be extracted from dried cyanobacteria under conditions that cleaved its bond to the phycobiliprotein.I remember becoming very excited about the prospect that an arbitrary protein could be fluorescently tagged in situ by genetically fusing it to the phycobiliprotein, then administering the chromophore, which I hoped would be able to cross membranes and get inside cells.Unfortunately, Glazer's lab then found out that the spontaneous reaction between the apoprotein and the chromophore produced the "wrong" product, whose fluorescence was red-shifted and five-fold lower than that of the native phycobiliprotein [1][2][3] .An enzyme from the cyanobacteria was required to insert the chromophore correctly into the apoprotein.This enzyme was a heterodimer of two gene products, so at least three cyanobacterial genes would have to be introduced into any other organism, not counting any gene products needed to synthesize the chromophore 4 .Meanwhile fluorescence imaging of the second messenger cAMP (cyclic adenosine 3',5'-monophosphate) had become one of my main research goals by 1988.I reasoned that the best way to create a fluorescent sensor to detect cAMP with the necessary affinity and selectivity inside cells would be to hijack a natural cAMP-binding protein.After much consideration of the various candidates known at the time, I chose cAMP-dependent protein kinase, now more commonly abbreviated PKA.PKA contains two types of

The spread of cancer from organ to organ (metastasis) is responsible for the vast majority of cancer deaths; however, most current anti-cancer drugs are designed to arrest or reverse tumor growth without directly addressing disease spread. It was recently discovered that tumor cell-secreted interleukin-6 (IL-6) and interleukin-8 (IL-8) synergize to enhance cancer metastasis in a cell-density dependent manner, and blockade of the IL-6 and IL-8 receptors (IL-6R and IL-8R) with a novel bispecific antibody, BS1, significantly reduced metastatic burden in multiple preclinical mouse models of cancer. Bispecific antibodies (BsAbs), which combine two different antigen-binding sites into one molecule, are a promising modality for drug development due to their enhanced avidity and dual targeting effects. However, while BsAbs have tremendous therapeutic potential, elucidating the mechanisms underlying their binding and inhibition will be critical for maximizing the efficacy of new BsAb treatments. Here, we describe a quantitative, computational model of the BS1 BsAb, exhibiting how modeling multivalent binding provides key insights into antibody affinity and avidity effects and can guide therapeutic design. We present detailed simulations of the monovalent and bivalent binding interactions between different antibody constructs and the IL-6 and IL-8 receptors to establish how antibody properties and system conditions impact the formation of binary (antibody-receptor) and ternary (receptor-antibody-receptor) complexes. Model results demonstrate how the balance of these complex types drives receptor inhibition, providing important and generalizable predictions for effective therapeutic design.

The vascular endothelial growth factor (VEGF) family of cytokines are key drivers of blood vessel growth and remodeling. These ligands act via multiple VEGF receptors (VEGFR) and co-receptors such as Neuropilin (NRP) expressed on endothelial cells. These membrane-associated receptors are not solely expressed on the cell surface, they move between the surface and intracellular locations, where they can function differently. The location of the receptor alters its ability to 'see' (access and bind to) its ligands, which regulates receptor activation; location also alters receptor exposure to subcellularly localized phosphatases, which regulates its deactivation. Thus, receptors in different subcellular locations initiate different signaling, both in terms of quantity and quality. Similarly, the local levels of co-expression of other receptors alters competition for ligands. Subcellular localization is controlled by intracellular trafficking processes, which thus control VEGFR activity; therefore, to understand VEGFR activity, we must understand receptor trafficking. Here, for the first time, we simultaneously quantify the trafficking of VEGFR1, VEGFR2, and NRP1 on the same cells-specifically human umbilical vein endothelial cells (HUVECs). We build a computational model describing the expression, interaction, and trafficking of these receptors, and use it to simulate cell culture experiments. We use new quantitative experimental data to parameterize the model, which then provides mechanistic insight into the trafficking and localization of this receptor network. We show that VEGFR2 and NRP1 trafficking is not the same on HUVECs as on non-human ECs; and we show that VEGFR1 trafficking is not the same as VEGFR2 trafficking, but rather is faster in both internalization and recycling. As a consequence, the VEGF receptors are not evenly distributed between the cell surface and intracellular locations, with a very low percentage of VEGFR1 being on the cell surface, and high levels of NRP1 on the cell surface. Our findings have implications both for the sensing of extracellular ligands and for the composition of signaling complexes at the cell surface versus inside the cell.

As we wrote previously [1], science benefits when we share not only our insights and discoveries but also the tools and approaches that we develop.This sharing improves reproducibility and reuse, and it enables others to build on our work.These tools and approaches are research accelerants and are the focus of 2 key sections in our journal: Methods and Software.Since its founding in 2005, PLOS Computational Biology has been the home of exceptional computational research and cutting-edge methodological advances.We publish papers that use computational techniques to generate new biological insight, and we publish papers that describe software or methods that many other researchers in the field can use independently to generate new biological insight.There is great merit in using existing techniques to advance biological understanding, and there is great merit in developing and sharing new techniques that will result in further advances.The field of Computational Biology supports both.In 2013, the journal introduced the Methods section and the Software section to encourage more researchers to publish methodological advancements.From the inception of these sections, there were dedicated Methods Editors and Software Editors on the Editorial Board responsible for the papers submitted to each section.Through the end of 2022, those editors have guided the publication of 647 Methods papers and 286 Software papers, from among the 1,790 and 646 manuscripts submitted, respectively.The standard of those papers has been high, with many excellent and impactful papers published, covering methods for microbiome data analysis [2], deep learning to predict molecular interactions [3], and quantitative analysis of live-cell imaging [4], and software tools ranging from Bayesian Evolutionary Analysis [5] to multiomics integration and feature selection [6], and genome assembly [7].At the outset, the concept of having separate Methods-specific and Software-specific editors was seen as beneficial to being able to set appropriate criteria for the review and publication of these papers, for example, making clear that new work that facilitated original research was publishable, even if the paper itself did not include original research findings.By creating and stewarding a clear standard for these papers, it has been possible to maintain and even increase the high standard of methods and software papers published in our journal.By all metrics, the Methods and Software sections have been a massive success and continue to grow.More than half of all submissions received since 2013 in these sections have come in the last 3 years (Fig 1).Our editors made a Herculean effort to meet this demand, but it became harder to keep up.Clearly, we needed to spread the handling of these papers across more people.

On March 30, 2021, a new code sharing policy was introduced at PLOS Computational Biology [1].This policy requires any code supporting a publication to be shared unless there are ethical or legal restrictions that prevent sharing.The policy was introduced in response to community desire for a stronger position on code sharing to reflect the fact that the majority of the community already voluntarily share code [2,3].This community-driven support for open science practices aligns well with the PLOS mission, and, therefore, the implementation of the new policy was a logical progression for the journal.The policy focuses on increasing code sAU : Pleasenotet haring as its primary aim, which, in turn, will support reproducibility, and so is not prescriptive to authors about how or where to share their code.The policy (https://journals.plos.org/ ploscompbiol/s/code-availability) allows authors to comply in ways which work for them.By the end of the first year of the policy, we expected to see an increase in code sharing rates (the percentage of published research articles that share code) without any negative impact on the publishing demographics or the author, editor, and journal staff experiences.This Editorial reports on the impact of policy over the first 12 months, provides a longitudinal view of code sharing in the journal since 2019, and articulates how this effort can move forward to enhance further sharing, reproducibility, and openness.

Biology was launched in 2005 as a journal driven by the computational biology research community and with the principle that "open access ensures not only that everything we publish is immediately freely available to anyone, anywhere in the world, but also that the contents of this journal can be redistributed and reused in ways that increase their value."[1] Delivering this important open vision has relied on the enthusiasm and integrity of the broader computational biology community as editors, as reviewers, and as dedicated authors submitting excellent work to the journal.This vision of openness, redistribution, and reuse is so essential because most scientific progress builds on prior efforts.This progress is possible because as scientists we share not just our research results, but also our methods and protocols and key research tools.Open sharing allows others to check and reproduce our observations, and to build on our work, giving it even more impact over time.This value of sharing is not just true in experimental work, where a key antibody or plasmid must be shared to enable scrutiny and further research; it is also true in computational biology, where computational code is a key reagent.Given how essential newly developed code can be to computational biology research we have been collaborating with the Editorial Board of PLOS Computational Biology and consulting with computational biology researchers to develop a new more-rigorous code policy that is intended to increase code sharing on publication of articles.Code sharing is not new to many of our authors, and in 2019 over 40% of research articles published in the journal reported sharing some code.[2] Assuming this percentage to be a baseline of code sharing behaviour, and after consulting with individual researchers about their code sharing practices, we surveyed our authors and others working in the computational biology field.The objective of this survey was to determine what proportion of articles have code associated with them, and what proportion of this code has not been openly shared in the past.We also aimed to better understand challenges researchers have to overcome to share their code, and how a stronger code sharing policy might affect their opinion of the journal.[3] The results indicated that around 70% of articles have code associated with them based on the cohort who completed the survey.However, for about a third of these articles the authors have not shared the code in the past.The reasons given for this discrepancy range from practical issues, such as not having enough time, to legal and ethical reasons, which is supported by previous research into code sharing.[4] Analysis of the survey data suggested around 5% of articles could not share code for legal and ethical reasons, and therefore more articles published in PLOS Computational Biology could share code than have to date.While there are legitimate restrictions on the sharing of some code, we know that availability of code assists with the reproducibility of research, and journal policies are effective

There has been much discussion in the scientific literature on a crisis of reproducibility in science [1,2].It has been reported that the percentage of studies that are reproducible is as low as 10% or less, depending on the discipline [3].This inability to reproduce scientific findings from a given paper has been attributed to a lack of clarity in the methods and inherent variability in the biological system being studied [4].Reproducibility in computational biology research is certainly a problem, yet perhaps a challenge that our field can uniquely tackle.A lack of reproducibility in computational biology research can be attributed to many factors, but incomplete or erroneous descriptions of the simulations (e.g., which software version was used), incomplete documentation on how to run simulations, or simply failing to post the relevant computer code needed to run a given simulation are common issues that occur.Many tools have emerged that we can leverage to make computational biology research more reproducible (e.g., http://co.mbine.org/and https://normsys.h-its.org/)and there exist articles that propose best practices, such as Ten Simple Rules for Reproducible Computational Research [5] or Ten Simple Rules for Writing and Sharing Computational Analyses in Jupyter Notebooks [6].

Scientific advances are frequently catalyzed by exploring the intersection of disciplines.From its inception, PLOS Computational Biology has published key insights that advance our understanding of biology and medicine-advances enabled by developments in computation and quantitative analyses.As an example, a recent initiative across the journals PLOS Medicine, PLOS ONE, and PLOS Computational Biology [1] resulted in a fantastic collection of research in Machine Learning in Health and Biomedicine.The breadth of research published in PLOS Computational Biology at this intersection of machine learning and health and biology was incredible, from the use of machine learning analysis to delineate biomarkers for soft tissue sarcomas [2] to the prediction of antibiotic resistance in Escherichia coli from pan-genome data [3].We're only beginning to see the power of machine learning applied to health and biology, with the hope of identifying patterns in the biological and clinical data that will lead to biomarkers of disease and the development of new clinical intervention strategies.As these data-driven strategies evolve and mature, they may also lead to a richer understanding of biological mechanisms, enabling models to predict the outcomes of scenarios and perturbations beyond the bounds of previous studies and data.Many more cross-journal initiatives are in the works, exploring how disparate disciplines can be brought together to tackle seemingly intransigent problems with unique perspectives.Just launched is a Targeted Anticancer Therapies and Precision Medicine call for papers jointly with PLOS ONE and PLOS Computational Biology [4].With nearly 10 million people dying from cancer in 2018 [5] and an increasing appreciation of the heterogeneity of the disease [6], there is an urgent need to develop targeted therapies that can be dosed, scheduled, delivered, and combined in ways that are specific to the patient.Computational approaches to this precision medicine challenge can serve as the common framework that integrates the disparate fields of expertise needed to understand the biological, pharmacological, and physiological complexity of the system.Computation can link, leverage, and amplify expertise in all the areas that are needed to understand biology and medicine: molecular dynamics, biochemistry, cell biology, human physiology, pharmacometrics, clinical practice, and more.These interdisciplinary links enable us to predict protein structure changes from gene variant data, to predict pharmacodynamics of associated drug compounds, to identify correlations in data, and simply to handle and process the vast amounts of data.As we do in these scientific ventures, so we do in managing the activity of PLOS Computational Biology.This journal, like the other PLOS community journals, is led by more than 160 practicing scientists with a variety of disciplinary backgrounds.All papers submitted to the journal are evaluated by multiple scientists through the peer review process and by teams of scientists at the editorial stage.It is with this integration of these different opinions of scientists from different subdisciplines of computational biology that we try to identify and support the

The splice isoforms of vascular endothelial growth A (VEGF) each have different affinities for the extracellular matrix (ECM) and the coreceptor NRP1, which leads to distinct vascular phenotypes in model systems expressing only a single VEGF isoform. ECM-immobilized VEGF can bind to and activate VEGF receptor 2 (VEGFR2) directly, with a different pattern of site-specific phosphorylation than diffusible VEGF. To date, the way in which ECM binding alters the distribution of isoforms of VEGF and of the related placental growth factor (PlGF) in the body and resulting angiogenic signaling is not well-understood. Here, we extend our previous validated cell-level computational model of VEGFR2 ligation, intracellular trafficking, and site-specific phosphorylation, which captured differences in signaling by soluble and immobilized VEGF, to a multi-scale whole-body framework. This computational systems pharmacology model captures the ability of the ECM to regulate isoform-specific growth factor distribution distinctly for VEGF and PlGF, and to buffer free VEGF and PlGF levels in tissue. We show that binding of immobilized growth factor to VEGF receptors, both on endothelial cells and soluble VEGFR1, is likely important to signaling in vivo. Additionally, our model predicts that VEGF isoform-specific properties lead to distinct profiles of VEGFR1 and VEGFR2 binding and VEGFR2 site-specific phosphorylation in vivo, mediated by Neuropilin-1. These predicted signaling changes mirror those observed in murine systems expressing single VEGF isoforms. Simulations predict that, contrary to the 'ligand-shifting hypothesis,' VEGF and PlGF do not compete for receptor binding at physiological concentrations, though PlGF is predicted to slightly increase VEGFR2 phosphorylation when over-expressed by 10-fold. These results are critical to design of appropriate therapeutic strategies to control VEGF availability and signaling in regenerative medicine applications.

Matrix-binding isoforms and non-matrix-binding isoforms of vascular endothelial growth factor (VEGF) are both capable of stimulating vascular remodeling, but the resulting blood vessel networks are structurally and functionally different. Here, we develop and validate a computational model of the binding of soluble and immobilized ligands to VEGF receptor 2 (VEGFR2), the endosomal trafficking of VEGFR2, and site-specific VEGFR2 tyrosine phosphorylation to study differences in induced signaling between these VEGF isoforms. In capturing essential features of VEGFR2 signaling and trafficking, our model suggests that VEGFR2 trafficking parameters are largely consistent across multiple endothelial cell lines. Simulations demonstrate distinct localization of VEGFR2 phosphorylated on Y1175 and Y1214. This is the first model to clearly show that differences in site-specific VEGFR2 activation when stimulated with immobilized VEGF compared to soluble VEGF can be accounted for by altered trafficking of VEGFR2 without an intrinsic difference in receptor activation. The model predicts that Neuropilin-1 can induce differences in the surface-to-internal distribution of VEGFR2. Simulations also show that ligated VEGFR2 and phosphorylated VEGFR2 levels diverge over time following stimulation. Using this model, we identify multiple key levers that alter how VEGF binding to VEGFR2 results in different coordinated patterns of multiple downstream signaling pathways. Specifically, simulations predict that VEGF immobilization, interactions with Neuropilin-1, perturbations of VEGFR2 trafficking, and changes in expression or activity of phosphatases acting on VEGFR2 all affect the magnitude, duration, and relative strength of VEGFR2 phosphorylation on tyrosines 1175 and 1214, and they do so predictably within our single consistent model framework.

The human APOBEC3G is an innate restriction factor that, in the absence of Vif, restricts HIV-1 replication by inducing excessive deamination of cytidine residues in nascent reverse transcripts and inhibiting reverse transcription and integration. To shed light on impact of A3G-Vif interactions on HIV replication, we developed a multi-scale computational system consisting of intracellular (single-cell), cellular and extracellular (multicellular) events by using ordinary differential equations. The single-cell model describes molecular-level events within individual cells (such as production and degradation of host and viral proteins, and assembly and release of new virions), whereas the multicellular model describes the viral dynamics and multiple cycles of infection within a population of cells. We estimated the model parameters either directly from previously published experimental data or by running simulations to find the optimum values. We validated our integrated model by reproducing the results of in vitro T cell culture experiments. Crucially, both downstream effects of A3G (hypermutation and reduction of viral burst size) were necessary to replicate the experimental results in silico. We also used the model to study anti-HIV capability of several possible therapeutic strategies including: an antibody to Vif; upregulation of A3G; and mutated forms of A3G. According to our simulations, A3G with a mutated Vif binding site is predicted to be significantly more effective than other molecules at the same dose. Ultimately, we performed sensitivity analysis to identify important model parameters. The results showed that the timing of particle formation and virus release had the highest impacts on HIV replication. The model also predicted that the degradation of A3G by Vif is not a crucial step in HIV pathogenesis.

Vascular endothelial growth factor (VEGF) is a potent cytokine that binds to specific receptors on the endothelial cells lining blood vessels. The signaling cascade triggered eventually leads to the formation of new capillaries, a process called angiogenesis. Distributions of VEGF receptors and VEGF ligands are therefore crucial determinants of angiogenic events and, to our knowledge, no quantification of abluminal vs. luminal receptors has been performed. We formulate a molecular-based compartment model to investigate the VEGF distribution in blood and tissue in humans and show that such quantification would lead to new insights on angiogenesis and VEGF-dependent diseases. Our multiscale model includes two major isoforms of VEGF (VEGF(121) and VEGF(165)), as well as their receptors (VEGFR1 and VEGFR2) and the non-signaling co-receptor neuropilin-1 (NRP1). VEGF can be transported between tissue and blood via transendothelial permeability and the lymphatics. VEGF receptors are located on both the luminal and abluminal sides of the endothelial cells. In this study, we analyze the effects of the VEGF receptor localization on the endothelial cells as well as of the lymphatic transport. We show that the VEGF distribution is affected by the luminal receptor density. We predict that the receptor signaling occurs mostly on the abluminal endothelial surface, assuming that VEGF is secreted by parenchymal cells. However, for a low abluminal but high luminal receptor density, VEGF binds predominantly to VEGFR1 on the abluminal surface and VEGFR2 on the luminal surface. Such findings would be pertinent to pathological conditions and therapies related to VEGF receptor imbalance and overexpression on the endothelial cells and will hopefully encourage experimental receptor quantification for both luminal and abluminal surfaces on endothelial cells.

Members of the vascular endothelial growth factor (VEGF) family of proteins are critical regulators of angiogenesis. VEGF concentration gradients are important for activation and chemotactic guidance of capillary sprouting, but measurement of these gradients in vivo is not currently possible. We have constructed a biophysically and molecularly detailed computational model to study microenvironmental transport of two isoforms of VEGF in rat extensor digitorum longus skeletal muscle under in vivo conditions. Using parameters based on experimental measurements, the model includes: VEGF secretion from muscle fibers; binding to the extracellular matrix; binding to and activation of endothelial cell surface VEGF receptors; and internalization. For 2-D cross sections of tissue, we analyzed predicted VEGF distributions, gradients, and receptor binding. Significant VEGF gradients (up to 12% change in VEGF concentration over 10 mum) were predicted in resting skeletal muscle with uniform VEGF secretion, due to non-uniform capillary distribution. These relative VEGF gradients were not sensitive to extracellular matrix composition, or to the overall VEGF expression level, but were dependent on VEGF receptor density and affinity, and internalization rate parameters. VEGF upregulation in a subset of fibers increased VEGF gradients, simulating transplantation of pro-angiogenic myoblasts, a possible therapy for ischemic diseases. The number and relative position of overexpressing fibers determined the VEGF gradients and distribution of VEGF receptor activation. With total VEGF expression level in the tissue unchanged, concentrating overexpression into a small number of adjacent fibers can increase the number of capillaries activated. The VEGF concentration gradients predicted for resting muscle (average 3% VEGF/10 mum) is sufficient for cellular sensing; the tip cell of a vessel sprout is approximately 50 mum long. The VEGF gradients also result in heterogeneity in the activation of blood vessel VEGF receptors. This first model of VEGF tissue transport and heterogeneity provides a platform for the design and evaluation of therapeutic approaches.

Angiogenesis (neovascularization) plays a crucial role in a variety of physiological and pathological conditions including cancer, cardiovascular disease, and wound healing. Vascular endothelial growth factor (VEGF) is a critical regulator of angiogenesis. Multiple VEGF receptors are expressed on endothelial cells, including signaling receptor tyrosine kinases (VEGFR1 and VEGFR2) and the nonsignaling co-receptor Neuropilin-1. Neuropilin-1 binds only the isoform of VEGF responsible for pathological angiogenesis (VEGF165), and is thus a potential target for inhibiting VEGF signaling. Using the first molecularly detailed computational model of VEGF and its receptors, we have shown previously that the VEGFR-Neuropilin interactions explain the observed differential effects of VEGF isoforms on VEGF signaling in vitro, and demonstrated potent VEGF inhibition by an antibody to Neuropilin-1 that does not block ligand binding but blocks subsequent receptor coupling. In the present study, we extend that computational model to simulation of in vivo VEGF transport and binding, and predict the in vivo efficacy of several Neuropilin-targeted therapies in inhibiting VEGF signaling: (a) blocking Neuropilin-1 expression; (b) blocking VEGF binding to Neuropilin-1; (c) blocking Neuropilin-VEGFR coupling. The model predicts that blockade of Neuropilin-VEGFR coupling is significantly more effective than other approaches in decreasing VEGF-VEGFR2 signaling. In addition, tumor types with different receptor expression levels respond differently to each of these treatments. In designing human therapeutics, the mechanism of attacking the target plays a significant role in the outcome: of the strategies tested here, drugs with similar properties to the Neuropilin-1 antibody are predicted to be most effective. The tumor type and the microenvironment of the target tissue are also significant in determining therapeutic efficacy of each of the treatments studied.