ŠĻą”±į > ž’
¤ ¦ ž’’’ ” ¢ £ ’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’’ģ„Į M ųæ Ū\ bjbjā=ā= "^ W W ēB / T ’’ ’’ ’’ l ī ī ī øm øm øm 8 šm ¤ n ü õč q ü
| . Ę| Ę| Ę| ¾ Ų ä ē ē ē ē ē ē ē $ ź «ģ P >ē q ī ģ ŗ ¾ ģ ģ >ē @ Ę| Ę| ō Æč @ @ @ ģ $ Ę| ī Ę| ~ā @ ģ ē @ Ź @
źŹ ¢ L ī JĻ Ę| q °5ņtÉ ¶[ øm H jĢ , JĻ 4 Åč 0 õč Ģ “ ūģ X č ūģ JĻ @ Ł Supplemental
In silico, agent behavior was governed by literature-based rule-sets obtained from peer-reviewed independent literature (Tables S1-S4). Derivation of rule-sets in any ABM is inherently subjective, although in their entirety should accurately reflect the current understanding in the field. This section outlines the criteria and methods employed to generate the rule-set, as well as how they were instituted within the Netlogo ADDIN EN.CITE Wilensky19992862862869U. WilenskyNorthwestern UniversityNetLogohttp://ccl.northwestern.edu/netlogo1999Evanston, ILCenter for Connected Learning and Computer-based Modeling[1] software program. Also included are tables of rules that governed interactions in silico (Tables S1-S4), which should be consulted for explanation of the ABM, and a simplified logic diagram of hASC agent decisions (Figure 13). Of helpful reference may be a prior model ADDIN EN.CITE Bailey200744944944917Bailey, A. M.Thorne, B. C.Peirce, S. M.Department of Biomedical Engineering, University of Virginia, PO Box 800759, Charlottesville, VA, 22908, USA, speirce@virginia.edu.Multi-cell Agent-based Simulation of the Microvasculature to Study the Dynamics of Circulating Inflammatory Cell TraffickingAnn Biomed EngAnnals of biomedical engineeringAnn Biomed Eng916-363562007Jun0090-6964 (Print)17436112http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=17436112 eng[2] that served as the basis for the development of this model (available for download on the Peirce-Cottler laboratory website; HYPERLINK "http://www.bme.virginia.edu/peirce" www.bme.virginia.edu/peirce).
hASC adhesion molecule expression (Table S1)
For experimental data to fit inclusion criteria for rule formulation:
Cells must be of human origin obtained either through liposuction or lipectomy procedures and be early to mid-passage (p=0-4) at point of assay.
Data obtained by testing fluorescence using flow cytometry.
Data obtained from independently published peer-reviewed scientific articles with a single exception: PSGL-1. This rule (0% probability of expression) was confirmed in-house (data unpublished).
Monocyte adhesion molecule expression (Table S1)
Monocyte-endothelium interactions during the adhesion cascade have been intensely studied and reported in the literature. In most instances, there were too many high-quality studies to cite individually for each adhesion molecule expression rule. Here, the reader was referred to a relevant review or text. Regardless of whether data indicated constitutive expression (100% probability of expression), the instituted rule within the simulation space was set at a 95% probability of positive expression, at baseline. This maintained a level of randomness in the system.
For experimental data to fit inclusion criteria for rule formulation:
Cells must be monocytes of human origin, isolated from whole blood.
Data obtained by testing fluorescence using flow cytometry.
Ideally, cells obtained from healthy patients.
There were exceptions. The rule for CD65 expression (25% probability of expression) was derived from data obtained examining peripheral blood monocytes isolated from healthy patients. In this study, only mean fluorescence intensity was reported making it necessary to normalize to CD11a expression (assumed to be constitutively expressed). Similar techniques were utilized for the generation of other adhesion molecule expression rules, when necessary.
Endothelial cellular adhesion molecule expression (Table S1)
For experimental data to fit inclusion criteria for rule formulation:
CAM expression determined using fluorescence and/or confocal microscopy.
Tissue source in order of preference: human endothelium obtained from healthy skeletal muscle microvasculature, human endothelium obtained from healthy coronary microvasculature, murine skeletal muscle microvasculature, murine coronary microvasculature.
CAM expression normalized to PECAM-1 (CD31) or other constitutively expressed endothelial cell marker, when necessary.
For rule formulation, it was not ideal to use data from in vitro studies if cell lines were immortalized or of umbilical vein origin because they are phenotypically different from skeletal muscle endothelium. In vitro data was also problematic if it was obtained from statically cultured endothelial cells. Judging endothelial cell CAM expression independent of fluid flow forces does not reproduce in vivo conditions that have been shown to be critical in CAM regulation/expression.
However, it was necessary to consult in vitro data or less-than-ideal in vivo data when other data was not available. For example, during rule formulation for PSGL-1 expression, a study was referenced where human umbilical vein endothelial cells (HUVECs), foreskin microvascular ECs, and atherosclerotic arteries were assayed using immunofluorescence. For previously stated reasons, this was a weaker rule and should be flagged for future parameterization or in vivo investigation. Similarly, the rule for P-selectin (CD62p) expression was formulated from data obtained from rat heart microvasculature and mouse skeletal muscle microvasculature.
In most instances, endothelial cell adhesion molecule expression was verified in other systems including retinal microvasculature, brain endothelium, or in vitro (protein or mRNA expression). These studies add additional validation to the formulation of the rule-set, but were not cited for reasons stated previously.
Tissue macrophage and smooth muscle cell adhesion molecule expression
There were no rules governing CAM expression of either tissue macrophages or smooth muscle cells. In the simulation space, tissue macrophages did not migrate or travel through the circulation; hence these rules were not necessary. Likewise, smooth muscle cells did not migrate or travel through the circulation. Instead, they served only to visually distinguish between vascular phenotypes. Future models should consider inclusion of additional rules to account for these cell types.
Chemokine secretion (Table S2)
Upon creation within the simulation space, each agent was assigned a percent probability of secreting each accounted-for chemokine (0-95% probability of secretion). Rules were accessed by every agent, at every time-step, and if probability was met (random number generator) then secretion was assigned as positive (conversely negative if probability was not met). If positive, secretion occurred instantaneously at that time-step (must be re-calculated at every time-step; instantaneously off).
For formulation of rules for baseline chemokine secretion, simplifications were necessary. We considered predominately cell phenotype. For example, during ABM construction the question was asked: Have microvascular endothelial cells been shown to secrete IL-1 bš š š,š ši n d e p e n d e n t o f e x t e r n a l s t i m u l i ? A n s w e r s t o t h e s e q u e s t i o n s w e r e s o u g h t f o r e v e r y c h e m o k i n e a n d e v e r y c e l l p h e n o t y p e p r e s e n t w i t h i n t h e m o d e l , e x c e p t f o r h A S C s a n d s m o o t h m u s c l e c e l l s ( n o p r o g r a m m e d c h e m o k i n e s e c r e t i o n a b i l i t y ) .
B a s i s f o r a n s w e r s t o these questions as being positive or negative were found in relevant literature and were cited in the rule-set. Data from literature was then synthesized to assign relative percent probabilities of secretion on a per cell basis. For example, parameters such chemokine concentration, length of secretion, etc. were used to group secretion rules according to a low to high probability of action (10% to 75%). These were not detailed rules; in most instances a general review of chemokine activity during ischemia could be referenced for verification.
Chemokine-induced chemokine secretion (Table S2)
In the simulation space, if an agent was exposed to a chemokine, it might have induced secretion of additional chemokines (assuming there is a rule and basis in the literature for this connection). During rule formulation, connections between exposure to specific chemokines and subsequent chemokine secretion were emphasized. This means that rules stated a chemokines ability to promote or inhibit behavior of specific cell types only.
Key criteria for rule formulation:
Evidence was found in relevant literature for connections between a specific chemokine and its ability to induce secretion of additional chemokines.
The effect of exposure to a specific chemokine was simplified to consider only whether it promoted, inhibited, or had no effect on additional chemokine secretion.
The ABM accounted for synergistic effects. For example, certain chemokines only worked in the presence of other chemokines.
Only data that quantified changes in protein expression or secretion were considered. In vivo, a change in mRNA expression does not necessarily translate to a change in protein secretion.
A rule for inhibition reflected experimental data that indicated blocking secretion, blocking a receptor, down-regulating a receptor, etc. Likewise a rule for promotion indicated changing affinities, upregulation of receptors, additional secretion, etc.
We developed a computational tool to investigate acute skeletal muscle ischemia. As such, data from literature were only used if effects were seen at less than eight hours, though preferably four hours, after initial stimulus.
Key criteria governing implementation in the simulation space:
An agent must see, or be exposed, to a stimulus prior to any induced secretion. These were parameterized rules dependent on cell phenotype and location. For example, an endothelial cell would see a stimulus if either a) surrounding EC in the same vessel segment was secreting the chemokine, b) a circulating monocyte that was secreting a chemokine has rolled or firmly adhered to that EC, or c) a tissue macrophage in the interstitium local to the vessel was secreting a chemokine. If these conditions were met, then a probability of 25% was assigned to whether that EC would see the stimulus. This was calculated for every cell within the simulation, at every time-step.
Once an agent had seen a stimulus, there was a probability of action associated with that cells response (i.e., promote or inhibit further secretion). This was set at 50% for all cell phenotypes (assuming there was a rule governing action).
Chemokine secretion occurred instantaneously if conditions were met, and after every time-step, chemokine secretion and chemokine exposure (see a stimulus) was turned off instantaneously (binary). In this way, the continued presence of activating stimuli was necessary for secretion to continue.
At every-time step, if additional activating stimuli were not seen or if secretion was not induced, cells would return to baseline secretion levels.
Rule-set: WSS-induced chemokine secretion (Table S2)
Wall shear stress-induced chemokine secretion was instituted within the simulation space in a manner similar to chemokine-induced chemokine secretion. Rules were simplified to consider only connections between changes in WSS and the resulting secretion. Unfortunately, the majority of available data in the literature was collected from statically cultured cells that experienced sudden changes in WSS. Ideally, data from cells that had experienced long-term culture under flow conditions before experiencing changes in WSS would have been assayed instead. Similarly, there were many studies that examined changes in CAM expression following induction of flow in vitro, but again, this was not sufficient in that it fails to distinguish between observed changes due to the onset of flow and those due to changes in the magnitude of flow.
Key criteria for understanding implementation in the simulation space:
Only endothelial cell agents undergo WSS-induced chemokine secretion. No other agent phenotype (cell type) was capable of sensing changes in WSS in silico.
Secretion was instantaneously on if correct stimuli were present (probability of action was met), as well as instantaneously off if fluid flow forces were removed.
The probability of secretion if changes in WSS were sensed was assigned to a relative scale between 10% and 80%, depending on the magnitude of change. This subjective handling considered magnitude of WSS, previous levels of WSS, length of secretion, and amount of chemokine secreted, as reported in the literature (assessed as: low, medium, high, and very high).
WSS-induced secretion occurred only during periods of changing WSS (e.g., at the onset of acute ischemic injury). This accounted for the phenomena of de-sensitization. Nitric oxide (NO) secretion was the exception (NO was secreted under baseline conditions in healthy skeletal muscle microvasculature, in silico).
When EC agents experienced changes in WSS, however, NO secretion changed well, similarly to other chemokines.
Transmigration-induced chemokine secretion (Table S2)
In silico, a transmigrating monocyte or hASC induced secretion of chemokines from local EC agents. During each event, the probability of secretion occurring was set between 10% and 40% for each surrounding EC (chemokine-specific). Rules were based on in vitro assays where monocytes and ECs were co-cultured together and secretion of chemokines into the media was quantified. Reported are synergistic effects; neither cell was able to secrete similar levels without cell-cell interactions. The probability of action was set on a relative scale, considering cell number, degree of interaction, concentration of secreted chemokine, and time-span of secretion reported in the literature. For these rules, hASCs were assumed to behave as monocytes.
WSS-induced change in adhesion molecule expression (Table S3)
Rules were similar to those that governed WSS-induced chemokine secretion. Again, only changes in WSS were considered relevant. In silico, expression of CD54 and CD106 by EC agents were affected, and only occurred at the onset of ischemic injury and times after (coinciding with abrupt change in experienced WSS). It was instituted by changing the baseline expression level of these CAMs to 75% probability of expression for CD54 and 30% probability of expression for CD106. It is these new baseline-expression levels that each EC agent defaulted to, at every time-step, if other additional activating stimuli were not seen. There was discrepancy in the literature on whether an increase in WSS induced changes CD106 expression. A high-confidence rule was difficult to formulate because of reasons stated earlier (statically cultured ECs are inherently different than ECs cultured under flow). Nonetheless, in this model an increase in WSS increased the probability of positive CD106 expression. However, changes in WSS also induced increased secretion of NO, which mitigated this response (NO inhibited CD106 expression in silico). Future in vivo experimentation should be done to formulate higher confidence rules, and the effects of NO on other CAM should be included in future models.
Integrin activation (Table S4)
Prior to transmigration in silico, a circulating cell generally had to proceed through a several events first. A circulating cell would first survey its local environment to assess whether binding ligands were present. Second, the combined influences of chemokine and hemodynamics would force interactions and facilitate further survey behavior. At this point, for firm adhesion and transmigration to proceed, integrin activation must occur (firm adhesion without integrin activation was possible, although rare). Physiologically, integrin activation in silico represented the conformational change of integrins present on circulating cells into a high affinity state (LFA-1, MAC-1, and VLA-4). Once this occurred, firm adhesion was more likely to occur. Once firmly adhered, a cell had to remain firmly adhered for 10 + 2 seconds before initiating transmigration (during that time, continued CAM and chemokine presence was necessary to maintain adhesion, at every time step). There was a probability of action assigned to integrin activation and transmigration, derived from independent experiments published in relevant, peer-reviewed literature. Conversely, the rule for the elapsed time necessary for transmigration to occur was scaled relative to the time required for other events, and this relative scaling was based on accepted dogma in the field. In that sense, it was arbitrary and not explicitly based on relevant literature.
Key for understanding rule-creation and in silico execution:
Integrin activation and transmigration occurred instantaneously.
Before a cell could proceed from firm adhesion to transmigration, it had to survey its environment and ensure that appropriate adhesion molecules and chemokines were still present. Most circulating cells remained firmly adhered for ten time steps, which was enough time for local environment conditions to change (hence necessity to re-survey).
There was a probability of action associated with integrin activation, firm adhesion, and transmigration as a function of chemokine presence, cell phenotype, cell history, hemodynamic forces, adhesion molecule expression, binding molecules present, and time.
A cell that was firmly adhered but failed to transmigrate was either released back into the free-stream (33%), initiated rolling along endothelium (33%), or remained firmly adhered for at least one additional time-step (33%). Continued firm adhesion and rolling was contingent upon appropriate external chemokines and adhesion molecules being present.
Integrin activation on hASCs was modeled after integrin activation on monocytes. This simplification was necessary because data was not available.
Rules were formulated from data taken from parallel plate flow chamber assays where adhesion molecule expression on substrates, WSS, and chemokine presence was tightly controlled. Data reporting percent of cells that firmly adhered under prescribed conditions was used to formulate in silico rules. For example, if in vitro experiment reported that when monocytes were flowed over the protein VCAM-1 under specific conditions, 45% of the cell population would firmly adhere, a rule was instituted where, if similar conditions were present, a single monocyte had a 45% probability of adhesion. In this way, each individual cell behaved according to its local environment and individual characteristics.
When experimental literature reported an increased ability of a circulating cell to firmly adhere following chemokine activation, this was instituted in silico as a promotion of integrin activation. However, data in vitro may have been the result of changes in the receptor, ligand, or other undetermined mechanism. This should not influence the outputs of the model, but should be considered before interpreting results. F o r e x a m p l e , i f s i m u l a t i o n s r e p o r t e d t h e i m p o r t a n c e o f S D F - 1 aš a c t i v a t i o n d u r i n g t r a f f i c k i n g , i t s h o u l d b e r e m e m b e r e d t h a t t h e A B M d o e s n o t d i f f e r e n t i a t e b e t w e e n e f f e c t s o n t h e r e c e p t o r o r e f f e c t s o n t h e l i g a n d .
O c c a s i o n a l l y , i t w a s n e c e s s a r y t o f o r m u l a t e r u les for integrin activation from data on neutrophil activation. However, it is likely that integrin behavior is conserved across leukocyte sub-populations. These rules, however, should be updated as additional experimental data becomes available.
hASC incorporation into ischemic tissue in silico (Table S5)
The inability of the model to reproduce hASC incorporation efficiencies of 3-10% into ischemic tissue (and 3-5x that of levels quantified in healthy tissue) led to a re-examination of the models underlying literature-based rules and, ultimately, the generation of a new hypothesis. As such, it is important to examine and justify: 1) why these incorporation levels were anticipated within the simulation space (i.e., do these data accurately reflect the current understanding in the literature?); 2) why hASC ability to undergo selectin-mediated rolling was investigated further, instead of other steps in the adhesion cascade; and 3) whether the parameterization and subsequent adjustment of other literature-based rules could have restored appropriate levels of incorporation, as well.
First, a survey of all known relevant literature reporting the use of hASCs in the treatment of ischemic injuries was performed (Table S5). In most instances, incorporation efficiencies were not reported or were not applicable to the simulations. For example, incorporation was not assayed at early time-points (incorporation may decrease as time increases), tissues were not perfused prior to harvesting to eliminate circulating cell contamination (cell counts may be artificially high), or it was difficult to confidently extrapolate tissue-level incorporation from cell counts in relatively few tissue sections (counts may be artificially low). For these reasons, much of the data surveyed was not suitable for comparison to in silico simulations. To complement the hASC data, studies reporting the use of other similar cell types were referenced. However, no consensus existed here, other than an agreement that incorporation following therapeutic delivery was variable and low. Bone marrow mononuclear cell and endothelial progenitor cell incorporation in models of hindlimb ischemia varied from 7-20% ADDIN EN.CITE ADDIN EN.CITE.DATA [7-9], and incorporation in a mouse model of myocardial infarction varied from 2.6-4.7% (dependent on delivery method; ADDIN EN.CITE ADDIN EN.CITE.DATA [10]). Similarly, incorporation efficiencies into injured tissue following myocardial infarction in human patients varied from 1.3-39% (dependent on cell phenotype and patient characteristics; ADDIN EN.CITE ADDIN EN.CITE.DATA [11]). These wide data ranges were all considered when formulating model expectations; it was believed that hASC incorporation efficiencies should be significantly greater than that seen in healthy limbs (3-5x greater) and represent approximately 10% of the total delivered cells. This was based on the following:
The most reliable hASC data was comparing cell counts in the ischemic limb to cell counts in the ischemic limb (Table S5; specificity). Tissue was assayed at 48 hours, which we anticipate would trend lower than tissues being assayed at less than 1 hour.
There may be overlapping ranges of incorporation efficiencies between hASCs and other related therapeutic cell types, in a variety of types of ischemic injury.
A yes/no binary response, with an order of magnitude difference between negative incorporation and positive incorporation was reasonable.
The simulation space in our ABM included only one microvascular network and not an entire tissue.
The authors recognize that these assumptions may have to be re-visited in future models; it was difficult to compare in silico incorporation levels to in vivo incorporation levels with absolute certainty (i.e., which independent data-set to verify simulation results with?). However, it is important to note that the inclusion of SBM-X in simulations significantly enhanced incorporation efficiencies over control runs where SBM-X did not exist. Thus, it was the relative increase in incorporation efficiency that prompted the formation of a new hypothesis concerning the rolling ability of hASCs.
Rolling during the adhesion cascade was investigated further because it was known that this was shown to be a bottleneck and rate-limiting step for the homing of neutrophils, and it was thought this could be the case for hASCs as well. Furthermore, it was interesting that hASCs do not express PSGL-1 (critical for leukocyte homing) and still traffic to ischemic tissues and confer a therapeutic benefit following i.v. delivery (Table S5). It was conceivable that inadequacies in the rules that govern firm adhesion and/or chemokine and cytokine secretion could have contributed to the low incorporation efficiencies, but this was believed to be a low likelihood. This was for the following reasons: 1) these rules were formulated with a higher confidence, based on a greater number of relevant published studies; and 2) the rules governing chemokine and cytokine behavior were simplified to consider only connections between cell-types (ignored potency, concentration, timing, etc.) and thus were high-confidence rules, as well. Nonetheless, systematic chemokine knockout experiments were performed in silico ADDIN EN.CITE Bailey200861961961917Bailey, A. M.Lawrence, M. B.Shang, H.Katz, A. J.Peirce, S. M.Adipose-derived stromal cells slowly roll and firmly adhere VCAM-1 under flow: Effects of SDF-1a pre-conditioning and therapeutic implications(In review). (In review).2008[12]. The possibility that more that one inadequacy within the rule-set could have theoretically restored expected incorporation efficiencies further emphasizes the importance of pairing all computational simulations with in vitro and/or in vivo experimentation ADDIN EN.CITE Thorne200745145145117Thorne, Bryan C.Bailey, Alexander M.Peirce, Shayn M.Combining experiments with multi-cell agent-based modeling to study biological tissue patterningBrief BioinformBrief Bioinformbbm0242007June 21, 2007http://bib.oxfordjournals.org/cgi/content/abstract/bbm024v1 10.1093/bib/bbm024[13].
Supplemental Tables
Adhesion moleculeCell
Positive Population
CitationLFA-1 (CD11a)hASC1% ADDIN EN.CITE ADDIN EN.CITE.DATA [3,4]Monocyte95%established; ADDIN EN.CITE ADDIN EN.CITE.DATA [5-7]EC0%assumedMAC-1 (CD11b/CD18)hASC27% ADDIN EN.CITE ADDIN EN.CITE.DATA [3,8]Monocyte95%established; ADDIN EN.CITE ADDIN EN.CITE.DATA [5-7]EC0%assumedCD15hASC3% ADDIN EN.CITE Festy200515315315317Festy, F.Hoareau, L.Bes-Houtmann, S.Pequin, A. M.Gonthier, M. P.Munstun, A.Hoarau, J. J.Cesari, M.Roche, R.Laboratoire de Biochimie et Genetique Moleculaire, EA-2526, Universite de La Reunion, Faculte des Sciences, 15 Avenue Rene Cassin, 97415, Saint Denis, France.Surface protein expression between human adipose tissue-derived stromal cells and mature adipocytesHistochem Cell BiolHistochem Cell Biol113-211242Adipocytes/cytology/*metabolismAdipose Tissue/cytology/*metabolismAdultAgedAntigens, CD/metabolismBiological Markers/metabolismCell SeparationCells, CulturedFemaleFlow CytometryHumansMembrane Proteins/*metabolismMiddle AgedOmentum/cytology/metabolismStromal Cells/cytology/*metabolismSubcutaneous Tissue/metabolism2005Aug16032396http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=16032396 [9]Monocyte90% ADDIN EN.CITE ADDIN EN.CITE.DATA [7]EC0%assumedVLA-4 (CD29/CD49d)hASC55% ADDIN EN.CITE ADDIN EN.CITE.DATA [3,10-12]Monocyte95%established; ADDIN EN.CITE ADDIN EN.CITE.DATA [6,7,13]EC0%assumed; ADDIN EN.CITE Hemler199028428428417Hemler, M. E.Elices, M. J.Parker, C.Takada, Y.Dana-Farber Cancer Institute, Laboratory of Immunochemistry, Division of Tumor Virology, Boston, MA 02115.Structure of the integrin VLA-4 and its cell-cell and cell-matrix adhesion functionsImmunol RevImmunol Rev45-65114Amino Acid SequenceAnimalsCell AdhesionCell Adhesion Molecules/metabolismEndothelium, Vascular/metabolismExtracellular Matrix/*metabolismHumansMolecular Sequence DataMolecular StructureReceptors, FibronectinReceptors, Immunologic/metabolismReceptors, Very Late Antigen/*physiologyResearch Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, P.H.S.T-Lymphocytes, Cytotoxic/metabolism1990Apr2142475http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=2142475 [14]CD34hASC52% ADDIN EN.CITE ADDIN EN.CITE.DATA [3,9,12,15]Monocyte0%assumedEC 77% ADDIN EN.CITE ADDIN EN.CITE.DATA [16,17]CD54 (ICAM-1)hASC27% ADDIN EN.CITE ADDIN EN.CITE.DATA [3,8]Monocyte25% ADDIN EN.CITE ADDIN EN.CITE.DATA [6,18,19]EC66% ADDIN EN.CITE ADDIN EN.CITE.DATA [16]CD62E (E-selectin)hASC2% ADDIN EN.CITE ADDIN EN.CITE.DATA [3,20]Monocyte0%assumed; ADDIN EN.CITE Elangbam199726526526517Elangbam, C. S.Qualls, C. W.Dahlgren, R. R.Cell adhesion molecules--updateVet PatholVet Pathol61--73341Animals, Cell Adhesion Molecules, Humans, Immunoglobulins, 91505511997Jan[5]EC6.5% ADDIN EN.CITE ADDIN EN.CITE.DATA [16]CD62L (L-selectin)hASC0% ADDIN EN.CITE ADDIN EN.CITE.DATA [4,20]Monocyte90% ADDIN EN.CITE ADDIN EN.CITE.DATA [21,22]EC0%assumedCD62P (P-selectin)hASC0% ADDIN EN.CITE Katz200581818117Katz, A. J.Tholpady, A.Tholpady, S. S.Shang, H.Ogle, R. C.Department of Plastic and Reconstructive Surgery, University of Virginia, Charlottesville, VA 22908, USA. ajk2f@virginia.eduCell surface and transcriptional characterization of human adipose-derived adherent stromal (hADAS) cellsStem CellsStem Cells412-23233ATP-Binding Cassette Transporters/geneticsAdipose Tissue/chemistry/ cytology/metabolismAdultAntigens, Differentiation/analysisAntigens, Surface/ analysisCell Differentiation/physiologyCells, CulturedFlow CytometryGene Expression ProfilingHumansMiddle AgedNeoplasm Proteins/geneticsOligonucleotide Array Sequence AnalysisResearch Support, N.I.H., ExtramuralResearch Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, P.H.S.Reverse Transcriptase Polymerase Chain ReactionStromal Cells/chemistry/ cytology/metabolismTelomerase/geneticsTranscription, Genetic/genetics2005Mar1066-5099 (Print)15749936[4]Monocyte0%assumed; ADDIN EN.CITE ADDIN EN.CITE.DATA [5,23]EC5% ADDIN EN.CITE ADDIN EN.CITE.DATA [16,24,25]CD65hASC29% ADDIN EN.CITE Festy200515315315317Festy, F.Hoareau, L.Bes-Houtmann, S.Pequin, A. M.Gonthier, M. P.Munstun, A.Hoarau, J. J.Cesari, M.Roche, R.Laboratoire de Biochimie et Genetique Moleculaire, EA-2526, Universite de La Reunion, Faculte des Sciences, 15 Avenue Rene Cassin, 97415, Saint Denis, France.Surface protein expression between human adipose tissue-derived stromal cells and mature adipocytesHistochem Cell BiolHistochem Cell Biol113-211242Adipocytes/cytology/*metabolismAdipose Tissue/cytology/*metabolismAdultAgedAntigens, CD/metabolismBiological Markers/metabolismCell SeparationCells, CulturedFemaleFlow CytometryHumansMembrane Proteins/*metabolismMiddle AgedOmentum/cytology/metabolismStromal Cells/cytology/*metabolismSubcutaneous Tissue/metabolism2005Aug16032396http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=16032396 [9]Monocyte25% ADDIN EN.CITE Stohlawetz199854154154117Stohlawetz, P.Hahn, P.Koller, M.Hauer, J.Resch, H.Smolen, J.Pietschmann, P.Department of Rheumatology, University of Vienna, and Hospital Barmherzige Schwestern, Austria.Immunophenotypic characteristics of monocytes in elderly subjectsScand J ImmunolScandinavian journal of immunologyScand J Immunol324-6483AdultAged/*physiologyAged, 80 and overAntigens, CD29/analysisAntigens, CD45/analysisCell AdhesionCell CountCell MovementCells, Cultured/immunologyEndothelium, Vascular/cytologyFemaleHumansImmunophenotypingIntercellular Adhesion Molecule-1/analysisMaleMonocytes/cytology/*immunologyReceptors, IgG/analysis1998Sep0300-9475 (Print)9743221http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=9743221 eng[26]EC0%assumedCD106 (VCAM-1) hASC15% ADDIN EN.CITE ADDIN EN.CITE.DATA [4,11]Monocyte0%assumed; ADDIN EN.CITE Elangbam199726526526517Elangbam, C. S.Qualls, C. W.Dahlgren, R. R.Cell adhesion molecules--updateVet PatholVet Pathol61--73341Animals, Cell Adhesion Molecules, Humans, Immunoglobulins, 91505511997Jan[5]EC3.5% ADDIN EN.CITE ADDIN EN.CITE.DATA [16]CD162 (PSGL-1)hASC0%Verified experimentallyMonocyte95% ADDIN EN.CITE Moore199527727727717Moore, K. L.Patel, K. D.Bruehl, R. E.Li, F.Johnson, D. A.Lichenstein, H. S.Cummings, R. D.Bainton, D. F.McEver, R. P.P-selectin glycoprotein ligand-1 mediates rolling of human neutrophils on P-selectinJ Cell BiolJ Cell Biol661--6711284Amino Acid Seque, Animals, Antibodies, Antibody Specificity, Base Sequence, Blotting, CHO Cells, Cell Adhesion, Cloning, Cricetinae, Friction, Immunoelectron, Leukocytes, Membrane Glycoproteins, Membrane Proteins, Microscopy, Microvilli, Molecular, Mole1995Feb[27]EC10% ADDIN EN.CITE ADDIN EN.CITE.DATA [28]SBM-XhASC75%Theorized
Table S1. Baseline CAM expression for monocytes, endothelial cells, and hASCs as implemented within the ABM.
Chemokine
SecretedActivatorResponderStimulus: ProbabilityCitationIL-1 bš B a s e l i n e M o n o c y t e / M Fš 4 0 % A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 2 9 - 3 3 ] C h e m o k i n e - i n d u c e d E C
E C
M o n o c y t e
M o n o c y t e
M Fš I L - 1 bš: p r o m o t e
T N F - aš: p r o m o t e
I L - 1 bš: p r o m o t e
M C P - 1 : p r o m o t e
T G F - bš: i n h i b i t A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 3 0 , 3 4 - 3 6 ] Dš in WSS-inducedECIncrease: promote ADDIN EN.CITE Jiang200420520520517Jiang, Z.Berceli, S. A.Pfahnl, C. L.Wu, L.Goldman, D.Tao, M.Kagayama, M.Matsukawa, A.Ozaki, C. K.University of Florida College of Medicine and the Malcom Randall Veterans Affairs Medical Center, Gainesville, 32610, USA.Wall shear modulation of cytokines in early vein graftsJ Vasc SurgJ Vasc Surg345-50402Animals*Blood Vessel ProsthesisCytokines/*immunologyHyperplasiaInterleukin-1/immunologyInterleukin-10/immunologyJugular Veins/*immunology/pathology/*physiopathologyMaleModels, AnimalRabbitsResearch Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, P.H.S.Shear StrengthTime Factors2004Aug15297832http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15297832 [37]IL-8BaselineEC
Monocyte20%
60% ADDIN EN.CITE ADDIN EN.CITE.DATA [32,38-40]Chemokine-inducedEC
EC
Monocyte
MonocyteIL-1 bš: p r o m o t e
T G F - bš: i n h i b i t
I L - 1 bš: p r o m o t e
T N F - aš: p r o m o t e A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 4 0 - 4 2 ] Dš i n W S S - i n d u c e d E C I n c r e a s e : p r o m o t e A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 4 3 , 4 4 ] T r a n s m i g r a t i o n E C 4 0 % A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 3 8 ] I L - 1 0 B a s e l i n e M o n o c y t e 2 0 % A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 3 2 , 4 5 ] C h e m o k i n e - i n d u c e d M o n o c y t e
M o n o c y t e
M Fš T N F - aš: p r o m o t e
I L - 1 0 : i n h i b i t
T G F - bš: p r o m o t e A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 4 5 - 4 7 ] T N F - aš B a s e l i n e M Fš
M o n o c y t e 3 0 %
6 0 % A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 3 0 - 3 2 ] C h e m o k i n e - i n d u c e d M o n o c y t e / M Fš
M Fš
M o n o c y t e I L - 1 0 : i n h i b i t
T G F - bš: i n h i b i t
T G F - bš: p r o m o t e A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 3 0 , 3 4 , 4 8 ] M C P - 1 B a s e l i n e E C 6 0 % A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 3 2 , 3 8 ] C h e m o k i n e - i n d u c e d E C
E C
E C
M Fš T N F - aš: p r o m o t e
I L - 1 bš: p r o m o t e
T G F - bš: i n h i b i t
T G F - bš: p r o m o t e A D D I N E N . C I T E A D D I N E N . C I T E . D A T A [ 4 9 , 5 0 ] Dš i n W S S - i n d u c e d E C I n c r e a s e : p r o m o t e A D D I N E N . C I T E <