Leveraging Algorithms to Improve Decision-Making Workflows for Genomic Data Access and Management

Studies on the ethics of automating clinical or research decision making using artificial intelligence and other algorithmic tools abound. Less attention has been paid, however, to the scope for, and ethics of, automating decision making within regulatory apparatuses governing the access, use, and exchange of data involving humans for research. In this article, we map how the binary logic flows and real-time capabilities of automated decision support (ADS) systems may be leveraged to accelerate one rate-limiting step in scientific discovery: data access management. We contend that improved auditability, consistency, and efficiency of the data access request process using ADS systems have the potential to yield fairer outcomes in requests for data largely sourced from biospecimens and biobanked samples. This procedural justice rationale reinforces a broader set of participant and data subject rights that data access committees (DACs) indirectly protect. DACs protect the rights of citizens to benefit from science by bringing researchers closer to the data they need to advance that science. DACs also protect the informational dignities of individuals and communities by ensuring the data being accessed are used in ways consistent with participant values. We discuss the development of the Global Alliance for Genomics and Health Data Use Ontology standard as a test case of ADS for genomic data access management specifically, and we synthesize relevant ethical, legal, and social challenges to its implementation in practice. We conclude with an agenda of future research needed to thoughtfully advance strategies for computational governance that endeavor to instill public trust in, and maximize the scientific value of, health-related human data across data types, environments, and user communities.

Empirical validation of an automated approach to data use oversight

The current paradigm for data use oversight of biomedical datasets is onerous, extending the timescale and resources needed to obtain access for secondary analyses, thus hindering scientific discovery. For a researcher to utilize a controlled-access dataset, a data access committee must review her research plans to determine whether they are consistent with the data use limitations (DULs) specified by the informed consent form. The newly created GA4GH data use ontology (DUO) holds the potential to streamline this process by making data use oversight computable. Here, we describe an open-source software platform, the Data Use Oversight System (DUOS), that connects with DUO terminology to enable automated data use oversight. We analyze dbGaP data acquired since 2006, finding an exponential increase in data access requests, which will not be sustainable with current manual oversight review. We perform an empirical evaluation of DUOS and DUO on selected datasets from the Broad Institute's data repository. We were able to structure 118/123 of the evaluated DULs (96%) and 52/52 (100%) of research proposals using DUO terminology, and we find that DUOS' automated data access adjudication in all cases agreed with the DAC manual review. This first empirical evaluation of the feasibility of automated data use oversight demonstrates comparable accuracy to human-based data access oversight in real-world data governance.

Stages in the Adoption of a Spatial Decision Support System for Reorganizing Service Delivery Regions

A prototype spatial decision support system (SDSS) was used by a task force appointed by the State of Iowa to redraw the boundary lines of education service delivery regions so that “the total number … is no fewer than four and no greater than twelve”. Details of this regionalization problem are described and key stages are identified in the task force's adoption and use of the SDSS to arrive at recommended solutions. The approach separates the policy task of setting regionalization objectives and criteria from the more technical task of locational analysis which involves searching for alternatives that meet the requirements of the policy criteria. It is concluded that the SDSS contributed significantly to meeting the basic objectives of the Iowa Legislature.

Retirement Migration Decision Making

This study examines the relative influence of a large number of characteristics of places on the migration decision making of a sample of persons in the age group in which persons will be considering retirement migration. An experimental design is used in which persons judge the desirability of migrating to hypothetical locations that vary in climate, terrain, closeness to family, and a number of other attributes. The findings show the relative importance given to 10 attributes of destinations and the distribution of preferred levels of each attribute when subjects rate the attractiveness of destinations described by multiple attributes.

Elucidation of the structural features of heparan sulfate important for interaction with the Hep-2 domain of fibronectin

The interaction of fibronectin with cell surface heparan sulfate proteoglycans is important biologically in inducing reorganization of the cytoskeleton and the assembly of focal adhesions. The major heparan sulfate-binding site in fibronectin, which is also implicated in these morphological events, is the COOH-terminal Hep-2 domain. We describe the first extensive study of the structural determinants required for the interaction between heparan sulfate/heparin and Hep-2. It is clear that, in heparan sulfate, there is a very prominent role for N-sulfate groups, as opposed to a relatively small apparent contribution from carboxyl groups. Furthermore, a minimal octasaccharide binding sequence appeared to contain at least two 2-O-sulfated iduronate residues, but no 6-O-sulfate groups. However, affinity was enhanced by the presence of 6-O-sulfates, and the interaction with Hep-2 also increased progressively with oligosaccharide size up to a maximum length of a tetradecasaccharide. This overall specificity is compatible with recent information on the structure of Hep-2 (Sharma, A., Askari, J. A., Humphries, M. J., Jones, E. Y., and Stuart, D. I. (1999) EMBO J. 18, 1468-1479) in which two separate, positively charged clusters, involving up to 11 basic amino acid residues (mostly arginines with their preferential ability to co-ordinate sulfate groups), could form a single extended binding site.

alpha4beta1 integrin-mediated adhesion of CD34+ cells from patients with chronic myeloid leukaemia: influence of IL-3

Interactions between integrins on haemopoietic progenitor cells and their stromal ligands have an important role in the control of haemopoiesis. Growth factors can modulate these interactions (so-called 'inside-out' signalling) resulting in changes in ligand binding activity. We have studied alpha4beta1 integrin-mediated adhesion to the H120 fragment of fibronectin (which contains the strongest alpha4beta1 binding site) in CD34+ cells from patients with chronic myeloid leukaemia (CML) and have determined the effect of IL-3 on the level of adhesion. Compared to normal CD34+ cells isolated from cord blood and peripheral blood progenitor harvests (mean of 61.4 +/- 14.9% of cells attached) the CML CD34+ cells showed reduced levels of adhesion (mean of 41.9 +/- 14.7%, P < 0.05). The effect of 10 ng/ml of IL-3 resulting in reduced adhesion of normal CD34+ cells at 30 min was absent in 6/7 patients with CML. Abnormalities of adhesion to fibronectin may thus be related to IL-3 pathways affected by BCR-ABL. These findings will have implications for understanding the dysregulation of growth and adhesion in CML.

Geographically masking health data to preserve confidentiality

The conventional approach to preserving the confidentiality of health records aggregates all records within a geographical area that has a population large enough to ensure prevention of disclosure. Though this approach normally protects the privacy of individuals, the use of such aggregated data limits the types of research one can conduct and makes it impossible to address many important health problems. In this paper we discuss the design and implementation of geographical masks that not only preserve the security of individual health records, but also support the investigation of questions that can be answered only with some knowledge about the location of health events. We describe several alternative methods of masking individual-level data, evaluate their performance, and discuss both the degree to which we can analyse masked data validly as well as the relative security of each approach, should anyone attempt to recover the identity of an individual from the masked data. We conclude that the geographical masks we describe, when appropriately used, protect the confidentiality of health records while permitting many important geographically-based analyses, but that further research is needed to determine how the power of tests for clustering or the strength of other associative relationships are adversely affected by the characteristics of different masks.

Geographically masking health data to preserve confidentiality

The conventional approach to preserving the confidentiality of health records aggregates all records within a geographical area that has a population large enough to ensure prevention of disclosure. Though this approach normally protects the privacy of individuals, the use of such aggregated data limits the types of research one can conduct and makes it impossible to address many important health problems. In this paper we discuss the design and implementation of geographical masks that not only preserve the security of individual health records, but also support the investigation of questions that can be answered only with some knowledge about the location of health events. We describe several alternative methods of masking individual-level data, evaluate their performance, and discuss both the degree to which we can analyse masked data validly as well as the relative security of each approach, should anyone attempt to recover the identity of an individual from the masked data. We conclude that the geographical masks we describe, when appropriately used, protect the confidentiality of health records while permitting many important geographically-based analyses, but that further research is needed to determine how the power of tests for clustering or the strength of other associative relationships are adversely affected by the characteristics of different masks.

Geographic information systems and public health: mapping the future

Images

Influence of interleukin-3 and other growth factors on alpha4beta1 integrin-mediated adhesion and migration of human hematopoietic progenitor cells

The mechanisms by which hematopoietic progenitor cells are normally anchored in stromal niches and yet can be mobilized by specific growth factors are poorly understood. It is likely, however, that integrins and their extracellular matrix (ECM) ligands play a key role in this process, and recent evidence suggests that integrin function is modulated by signals originating from activated growth factor receptors. We have now examined this further by studying the role of growth factors on alpha4beta1 integrin-mediated adhesion of human CD34+ hematopoietic progenitor cells to specific recombinant fibronectin fragments coated onto tissue culture dishes. Cells were prepared from cord blood and peripheral blood harvests. During a 30-minute adhesion assay a mean of 74% of CD34 cells attached to the so-called H120 fragment of fibronectin, which contains the strongest alpha4beta1 integrin-binding sequence. The level of cell adhesion was significantly reduced by low concentrations of interleukin-3 (IL-3) (2.5 to 10 ng/mL), whereas stem cell factor (SCF) and granulocyte colony-stimulating factor (G-CSF) at these concentrations did not affect adherence of the cells. Migratory behavior of CD34 cells was examined using fibronectin fragments adsorbed onto a Transwell filter. The H120 fragment supported much higher levels of cell migration than the H0 fragment of fibronectin, which contains a weak alpha4beta1 integrin binding sequence. Over a 16-hour assay, migration of peripheral blood progenitor cells was increased slightly by SCF and by G-CSF. However, a marked stimulation was observed with IL-3, which significantly increased migration. Similar effects were noted with cord blood cells, although a small proportion of cells were able to migrate in the absence of growth factors. These results indicate that there is a highly selective and functional link between the alpha4beta1 integrin and IL-3/IL-3-receptor that could affect the position of stem and progenitor cells in the marrow stroma and influence their growth and development.

The interaction of the transforming growth factor-betas with heparin/heparan sulfate is isoform-specific

We have undertaken a comparative study of the interaction of the three mammalian transforming growth factor-betas (TGF-beta) with heparin and heparan sulfate. TGF-beta1 and -beta2, but not -beta3, bind to heparin and the highly sulfated liver heparan sulfate. These polysaccharides potentiate the biological activity of TGF-beta1 (but not the other isoforms), whereas a low sulfated mucosal heparan sulfate fails to do so. Potentiation is due to antagonism of the binding and inactivation of TGF-beta1 by alpha2-macroglobulin, rather than by modulation of growth factor-receptor interactions. TGF-beta2.alpha2-macroglobulin complexes are more refractory to heparin/heparan sulfate, and those involving TGF-beta3 cannot be affected. Comparison of the amino acid sequences of the TGF-beta isoforms strongly implicates the basic amino acid residue at position 26 of each monomer as being a vital binding determinant. A model is proposed in which polysaccharide binding occurs at two distinct sites on the TGF-beta dimer. Interaction with heparin and liver heparan sulfate may be most effective because of the ability of the dimer to co-operatively engage two specific sulfated binding sequences, separated by a distance of approximately seven disaccharides, within the same chain.

The spatial relationship between infant mortality and birth defect rates in a U.S. city

The spatial patterns of infant mortality and birth defect rates in the Des Moines, Iowa, urban region are described as a contoured surface based on the punctual kriging of address-matched vital statistics records from The Iowa Department of Public Health. Areas defined as having high rates are shown to be sensitive to the size of the spatial filtering units. There is no correlation between infant mortality and birth defect rates in the region. The significance of areas with high rates is determined by a Monte Carlo simulation procedure. One area of high infant mortality is found in the region, which contrasts with many smaller areas with high birth defect rates in the region. The observed birth defect rate pattern is consistent with the hypothesis that each birth was equally likely to result in a birth defect, while the infant mortality pattern is unlikely to be the result of such an equal likelihood process.

Exploratory spatial analysis of birth defect rates in an urban population

Are birth defect rates unusually high in particular urban localities? The answer requires that the birth defect rate for which 'significance' is claimed be adjusted for the variable population sizes of each area for which the rate is computed and for the spatial dependence of rates based on shared observations between neighboring areas. By address-matching birth and birth defect records to a digital road map, we are able to compute local birth defect rates at regular grid locations by dividing the number of birth defects that occurred in the geographical vicinity of a grid location by the total number of births in the same vicinity. We test for significance, at regular spatial intervals, against the null hypothesis that the observed rate at any locality could reasonably have arisen by chance alone, given the underlying geographical variation in births. Significance is determined by using Monte Carlo simulations, where each birth location has an identical probability of being a defect. From 1000 simulations, a statistical distribution of the birth defect rate for each grid location is determined. The proportion of the simulated birth defect rates that are less than the observed rate at any grid location is the probability that the observed rate is significant. These probabilities, portrayed as isarithmic maps, show areas that have significantly high birth defect rates. Our results show birth defect rates for the period 1983 to 1990 in Des Moines, Iowa, U.S.A.

Phenotypic heterogeneity in breast fibroblasts: functional anomaly in fibroblasts from histologically normal tissue adjacent to carcinoma

Histologically normal breast tissue was obtained from women undergoing surgery for benign breast lesions (n = 12) and mammary carcinomas (n = 15). Four fibroblast subpopulations (FI, FII, FIII and FIV) were isolated from these specimens by differential digestion and centrifugation. FI cells were the first to be released from the tissue digest and consequently assumed to be derived from the interlobular stroma; FIV fibroblasts were tightly associated with the epithelial organoids and are therefore believed to be of intralobular origin. These cells were characterised in terms of their migratory phenotype (classified as either foetal- or adult-like) and the production of motility factors according to previously described techniques. FI fibroblasts obtained from patients with benign breast lesions displayed a foetal migratory phenotype (10/11) and secreted detectable quantities of motility factors (11/11). In contrast, none of the FIV fibroblasts (0/10) obtained from these same patients displayed a foetal-like migratory phenotype or secreted motility factors. In the case of fibroblasts obtained from cancer patients, both FI (13/13) and FIV (13/13) fibroblasts displayed a foetal-like migratory phenotype and secreted motility factors. Fibroblasts were also derived from skin (n = 12) and breast fat tissue (n = 4) of certain patients. In agreement with our previously published observations, skin fibroblasts obtained from non-cancer and cancer patients also differed in terms of their migratory behaviour: none of the skin fibroblast lines (0/5) obtained from non-cancer patients were foetal-like, compared to 3/7 lines from cancer patients. All fat-derived fibroblasts (1 non-cancer and 3 cancer patients) were also foetal-like. Our results indicate (i) functional heterogeneity between FI and FIV fibroblasts of normal breast, and (ii) the presence of functionally aberrant (i.e., foetal-like) FIV fibroblasts in histologically normal breast tissue adjacent to a carcinoma.

Heterogeneity amongst fibroblasts in the production of migration stimulating factor (MSF): implications for cancer pathogenesis

Fetal skin fibroblasts migrate into 3D collagen gels to a significantly greater extent than do adult cells. This enhanced motility of fetal fibroblasts appears to result from the production of a "migration stimulating factor" (MSF) which is not made by their normal adult counterparts. Adult skin fibroblasts retain responsiveness to MSF and cells exposed to this factor achieve the elevated levels of migration characteristic of fetal cells. MSF has been purified to homogeneity, has an apparent molecular mass of 70 kD and has been further characterized in terms of a number of biochemical parameters. Studies concerned with the mechanism of action of MSF indicate that it stimulates the production of a high molecular weight class of hyaluronic acid (HA). Concurrent exposure of cells to Streptomyces hyaluronidase blocks the stimulation of adult fibroblast migration by MSF. In a related series of experiments, we have shown that TGF-beta inhibits the effects of MSF on both cell migration and HA production. Taken together, these data suggest that the stimulation of fibroblast migration by MSF is dependent upon (and may directly result from) a primary induction of HA synthesis. We have previously reported that skin fibroblasts obtained from patients with sporadic and familial breast cancer, as well as the unaffected first-degree relatives of familial breast cancer patients, commonly display a fetal-like migratory phenotype. Subsequent work has indicated that (a) these fetal-like cells also produce MSF, and (b) detectable levels of MSF are present in the serum of sporadic breast cancer patients both prior to and following surgical resection of the primary tumor mass. On the basis of these and related observations, we have put forward an hypothesis suggesting that the disruption in normal epithelial-mesenchymal interactions caused by the persistent production of MSF by fibroblasts in the adult may contribute directly to the pathogenesis of an epithelial cancer. The demonstration of aberrant fibroblasts in sporadic cancer patients (both in our own and independent studies) is not consistent with the "germ-line genetic lesion" model commonly invoked to account for the presence of such cells in patients with hereditary cancer syndromes.(ABSTRACT TRUNCATED AT 400 WORDS)

Mechanism of action of the migration stimulating factor produced by fetal and cancer patient fibroblasts: effect on hyaluronic and synthesis

We have previously demonstrated that confluent fetal fibroblasts migrate into three-dimensional collagen gels to a significantly greater extent than their normal adult counterparts. Recent studies have revealed that this behavioral difference results from the secretion by fetal fibroblasts of a soluble migration-stimulating factor (MSF) which acts on these cells in an autocrine fashion. Adult fibroblasts do not produce MSF but remain responsive to it. Skin fibroblasts from cancer patients resemble fetal fibroblasts (rather than normal adult cells) with respect to their migratory behavior on collagen gels and continued production of MSF. This communication is concerned with elucidating the biochemical basis of MSF activity. Data are presented indicating that a) hyaluronic acid is required for the elevated migratory activity displayed by confluent fetal and breast cancer patient skin fibroblast; b) adult fibroblasts exhibit a bell-shaped dose-response to MSF, with maximal stimulation of migration observed at a concentration of 10 ng/ml; c) the migratory activity of adult fibroblasts pre-incubated with MSF remains high in the absence of additional factor: and d) MSF affects both the quantity and size class distribution of hyaluronic acid synthesized by adult fibroblasts. We have previously speculated that the persistent fetal-like fibroblasts of breast cancer patients play a direct role in disease pathogenesis by perturbing normal epithelial-mesenchymal interactions. The observations reported here suggest that MSF-induced alterations in hyaluronic acid synthesis may contribute to the molecular basis of such perturbations.

Purification of the migration stimulating factor produced by fetal and breast cancer patient fibroblasts

We have previously shown that (i) human skin fibroblasts of fetal and adult origin display distinctive migratory phenotypes, (ii) this difference in cell behavior results from the production of a soluble "migration stimulating factor" (MSF) by fetal cells, and (iii) skin fibroblasts from breast cancer patients commonly resemble fetal fibroblasts both in migratory phenotype and in production of MSF. Data are now presented indicating that MSF present in the conditioned medium of fetal and cancer patient fibroblasts is precipitated at 10% saturation ammonium sulfate and binds to heparin and cation-exchange resins. Based on this information, we have devised a scheme for the purification of MSF involving the sequential application of ammonium sulfate precipitation, heparin affinity, gel filtration, and reverse-phase chromatography. Purified MSF has an estimated molecular mass of 70 kDa; amino acid analysis reveals a relatively high level of proline (13.34 residues per 100). Our results further suggest that skin fibroblasts from breast cancer patients produce an additional factor with migration stimulating activity; this factor is precipitated at higher concentrations of ammonium sulfate and binds to anion-exchange resins. We have previously discussed the possible direct involvement of fetal-like fibroblasts in cancer pathogenesis. The availability of MSF obtained from cancer patient fibroblasts provides a potential means with which to examine the complex cellular interactions contributing to this process as well as develop a screening regime for identifying individuals at elevated risk of developing cancer.

Foetal and cancer patient fibroblasts produce an autocrine migration-stimulating factor not made by normal adult cells

We have previously reported that (1) the migration of foetal and adult fibroblasts into three-dimensional collagen matrices is differentially affected by cell density, and (2) skin fibroblasts from cancer patients commonly display a foetal-like mode of migratory behaviour. Data presented here indicate that differences in the migration of these cell types are particularly apparent in cultures plated at high density (i.e. at cell confluence); under these conditions, foetal fibroblasts and the foetal-like fibroblasts of cancer patients migrate into the three-dimensional collagen matrix to a significantly greater extent than do normal adult cells. In this initial study concerned with the biochemical basis of these observations, we report that medium conditioned by either foetal or cancer patient fibroblasts stimulates the migration of confluent adult cells. This stimulation of migration is specific to confluent cells, as the migration of subconfluent adult fibroblasts is unaffected by these conditioned media. Gel filtration chromatography of foetal fibroblast-conditioned medium indicates that migration-stimulating activity is recovered in a single peak with an apparent molecular mass in the range of 50–60 (X 10(3]. The active migration stimulating factor (MSF) in both foetal and cancer patient fibroblast-conditioned media appears to be a protein stable at acid pH, but inactivated by heat, alkaline pH and reductive alkylation. MSF produced by foetal and cancer patient fibroblasts is presumably responsible for the characteristically elevated levels of migration displayed by these cells in confluent culture, thereby suggesting an autocrine mode of action for this factor.(ABSTRACT TRUNCATED AT 250 WORDS)

Fibroblasts from cancer patients display a mixture of both foetal and adult-like phenotypic characteristics

We have previously reported that foetal and adult fibroblasts display distinctive migratory phenotypes when cultured on three-dimensional collagen gels. Both skin and tumour-derived fibroblasts from a significant proportion of patients with breast cancer were subsequently observed to display foetal-like migratory behaviour. In the accompanying paper concerned with the biochemical basis of these observations, we presented evidence that foetal fibroblasts and the foetal-like fibroblasts of cancer patients produce a soluble migration-stimulating factor (MSF) not made by normal adult cells. Data are presented here indicating that: (1) the spontaneous foetal-to-adult transition in migratory phenotype that foetal fibroblasts undergo after approximately 50-55 population doublings in vitro is correlated with a cessation of MSF production; (2) breast cancer patient fibroblasts do not undergo such a phenotypic transition and continue to produce MSF for their entire in vitro lifespan. These foetal-like cancer patient fibroblasts do, however, resemble normal adult cells by a number of other criteria, including population doubling potential, enhanced migration in the presence of serum compared to platelet-poor plasma, saturation cell density and morphology in confluent culture. These data indicate that the fibroblasts of breast cancer patients express a mixture of both foetal and adult phenotypic characteristics. Such a finding is consistent with published information indicating that foetal-to-adult transitions in various fibroblast phenotypic characteristics occur in a temporally disparate fashion during normal development, and further imply that cancer patient fibroblasts have undergone only certain of these transitions.

Improving the geographical accessibility of health care in rural areas: a Nigerian case study

The paper addresses problems of geographical accessibility of health care in rural areas of Nigeria. It provides analyses of the location, distribution and accessibility of government-provided health care facilities to people and presents a framework for measuring improvements in accessibility and for assessing the efficiency of decisions about location of new facilities. It shows that while accessibility in the study area improved between 1979 and 1982 through the establishment of more dispensaries and maternity and child-welfare centres, the relative efficiency of locations has remained low. It identifies alternate locations for the new facilities introduced in the 1979-1982 period that could have increased the utilization of maternal and child health centres by an estimated 12% and the utilization of dispensaries by 16%.

Skin fibroblasts obtained from cancer patients display foetal-like migratory behaviour on collagen gels

When plated on the surface of collagen gel substrata, all types of fibroblasts rapidly begin to migrate down into the three-dimensional collagen matrix. We have previously demonstrated that normal (adult and foreskin), foetal and transformed fibroblasts may be distinguished from each other by virtue of their differential migratory response to changes in cell density. The effects of cell density on fibroblast migration into the gel may be expressed by a single numerical value, the ‘cell density migration index’ (CDMI). We now present evidence that ostensibly normal skin fibroblasts obtained from the majority of patients we examined with carcinoma of the breast, malignant melanoma, familial polyposis coli, retinoblastoma and Wilms' tumours display aberrant CDMI values falling within the foetal range. Skin fibroblasts obtained from the majority of patients examined with genetic or chronic diseases (e.g. rheumatoid arthritis, Duchenne muscular dystrophy) displayed CDMI values falling within the normal range.

Adult, foetal and transformed fibroblasts display different migratory phenotypes on collagen gels: evidence for an isoformic transition during foetal development

Data are presented indicating that the migration of fibroblasts into three-dimensional collagen gels is affected by cell density. We have defined a ‘cell density migration index’ (CDMI) to express this behavioural response in quantitative terms. The results of a survey of 77 different cell types indicate that the CDMI values expressed by normal adult skin fibroblasts and transformed cell lines fall into two distinct, non-overlapping groups. Measurement of the CDMI therefore provides an additional means of distinguishing between normal and transformed cells and may be used in conjunction with other commonly recognized criteria (e.g. anchorage-independent growth) to assess expression of a transformed phenotype in vitro. It is of interest to note that the CDMI values expressed by foetal cells define a group lying intermediate between normal and transformed cells. Both uncloned and cloned foetal cells have been observed to undergo a stable transition to expression of CDMI values characteristic of adult cells when followed throughout the duration of their in vitro lifespan. In addition to providing a novel means of distinguishing between normal adult and foetal cells, our results suggest that foetal fibroblasts undergo an ‘isoformic’ transition at some point in their developmental history, which is manifest in vitro by the expression of an adult CDMI.

The use of three-dimensional collagen gels for the study of tumour cell invasion in vitro: experimental parameters influencing cell migration into the gel matrix

We have been using three-dimensional gels of native collagen fibres as a substratum for the study of cell migratory behaviour. The objective of this work has been to develop an experimental model system for the study of tumour cell invasion. Data are presented in this communication concerning the effects of a number of experimental parameters on the migration of human foreskin fibroblasts (normal) and a Syrian hamster melanoma cell line (tumour) into three-dimensional gels of native collagen fibres. We find that the migration of both these cell types into the gel matrix is affected by the concentration of collagen in the gel, as well as by the initial density of cells plated onto the gel and the density of cells in the stock culture dishes. These results are discussed in terms of the control of cell migration into the gel and also demonstrate the potential utility of this model system for the study of tumour cell invasion.