In Vivo Confocal Microscopy 1 Year after Autologous Cultured Limbal Stem Cell Grafts.

PURPOSE: To correlate clinical, impression cytologic, and in vivo confocal microscopy findings on the corneal surface after cultured limbal stem cell transplantation. DESIGN: Prospective, interventional, noncomparative, masked case series. PARTICIPANTS: Thirteen patients with limbal stem cell deficiency after unilateral (9 eyes) or bilateral (2 eyes) chemical burn, liquid nitrogen injury (1 eye), or herpes simplex virus infection (1 eye). METHODS: Limbal cells were harvested from healthy or less affected eyes, cultured on 3T3 cells and fibrin glue, and transplanted to the patient's injured eye. Patients underwent clinical examination and impression cytologic examination of the central cornea before and 1 year after intervention. In vivo confocal microscopy scans were obtained in all corneal quadrants after 1 year. The interexamination agreement was established by calculation of the Cohen's κ coefficient. MAIN OUTCOME MEASURES: Results of surgery were assessed considering clinical signs (successful: restoration of transparent, avascular, and stable corneal epithelium without neovascularization in central corneal surface; partially successful: recurrence of superficial neovascularization; failed: recurrent epithelial defects, pannus, and inflammation), phenotype of cells covering the corneal surface (conjunctivalized corneal surface: cytokeratin 12 [cK12]-negative and mucin 1 [MUC1]-positive cells; mixed epithelium: cK12-positive and MUC1-positive cells; corneal epithelium: cK12-positive and MUC1-negative cells), and cell morphologic features (corneal epithelium: multilayered polygonal and flat cells with hyperreflective nuclei; conjunctival epithelium: stratified cuboidal or polygonal cells, hyperreflective cytoplasm, and barely defined borders; epithelial transition: transition of epithelial cells from the cornea to the conjunctiva over the corneal surface). RESULTS: We found a moderate to substantial degree of concordance between confocal microscopy and clinical evaluation (κ = 0.768) and between confocal microscopy and impression cytologic analysis (κ = 0.629). Confocal microscopy showed that 46.2% of patients exhibited corneal epithelium in the central and peripheral cornea, 30.8% showed an irregular mixed corneal and conjunctival epithelium, and 23.0% showed conjunctival epithelium. Palisades of Vogt were absent in all (100.0%) patients, and the cornea-conjunctiva epithelial transition localized approximately 1 mm internally on the cornea. CONCLUSIONS: Confocal microscopy provides objective measures of the corneal epithelium and may significantly improve the evaluation of outcomes after cultured limbal stem cell graft.

Measurement of mechanical tractions exerted by cells in three-dimensional matrices.

Quantitative measurements of cell-generated forces have heretofore required that cells be cultured on two-dimensional substrates. We describe a technique to quantitatively measure three-dimensional traction forces exerted by cells fully encapsulated in well-defined elastic hydrogel matrices. Using this approach we measured traction forces for several cell types in various contexts and revealed patterns of force generation attributable to morphologically distinct regions of cells as they extend into the surrounding matrix.

Human hair follicle dermal sheath and papilla cells support keratinocyte growth in monolayer coculture.

Traditional skin grafting techniques are effective but limited methods of skin replacement. Autologous transplantation of rapidly cultured keratinocytes is successful for epidermal regeneration, but the current gold-standard technique requires mouse fibroblast feeders and serum-rich media, with serum-free systems and dermal fibroblast (DF) feeders performing relatively poorly. Here, we investigated the capacity of human hair follicle dermal cells to act as alternative supports for keratinocyte growth. Dermal papilla (DP) dermal sheath (DS), DF and 3T3 cells were used as inactivated feeder cells for human keratinocyte coculture. Under conditions favouring dermal cells, proliferation of keratinocytes in the presence of either DS or DP cells was significantly enhanced compared with DF cells, at levels comparable to keratinocytes cultured under gold-standard conditions. Secreted protein acidic and rich in cysteine (SPARC) expression increased DS and DP cells relative to DFs; however, further experiments did not demonstrate a role in keratinocyte support.

Association of SET domain and myotubularin-related proteins modulates growth control.

Several proteins that contribute to epigenetic mechanisms of gene regulation contain a characteristic motif of unknown function called the SET (Suvar3-9, Enhancer-of-zeste, Trithorax) domain. We have demonstrated that SET domains mediate highly conserved interactions with a specific family of proteins that display similarity with dual-specificity phosphatases (dsPTPases). These include myotubularin, the gene of which is mutated in a subset of patients with X-linked myotubular myopathy, and Sbf1, a newly isolated homologue of myotubularin. In contrast with myotubularin, Sbf1 lacks a functional catalytic domain which dephosphorylates phospho-tyrosine and serine-containing peptides in vitro. Competitive interference of endogenous SET domain-dsPTPase interactions by forced expression of Sbf1 induced oncogenic transformation of NIH 3T3 fibroblasts and impaired the in vitro differentiation of C2 myoblast cells. We conclude that myotubularin-type phosphatases link SET-domain containing components of the epigenetic regulatory machinery with signalling pathways involved in growth and differentiation.

Assembly of focal adhesions: progress, paradigms, and portents.

Receptor-mediated assembly of an adhesion plaque occurs through an ordered series of steps, and intermediate assemblies can be identified. The recent demonstration of some of these partial reactions in permeabilized cells predicts that cell-free reconstitution of adhesion plaque assembly is an attainable goal. Newly discovered cryptic actin-binding sites in vinculin and ezrin, two proteins recruited to adhesion sites, suggest that actin-binding proteins are targets for the signals generated by adhesion receptors.

mDia mediates Rho-regulated formation and orientation of stable microtubules.

Rho-GTPase stabilizes microtubules that are oriented towards the leading edge in serum-starved 3T3 fibroblasts through an unknown mechanism. We used a Rho-effector domain screen to identify mDia as a downstream Rho effector involved in microtubule stabilization. Constitutively active mDia or activation of endogenous mDia with the mDia-autoinhibitory domain stimulated the formation of stable microtubules that were capped and oriented towards the wound edge. mDia co-localized with stable microtubules when overexpressed and associated with microtubules in vitro. Rho kinase was not necessary for the formation of stable microtubules. Our results show that mDia is sufficient to generate and orient stable microtubules, and indicate that Dia-related formins are part of a conserved pathway that regulates the dynamics of microtubule ends.

Oncogenes in Ras signalling pathway dictate host-cell permissiveness to herpes simplex virus 1.

The importance of herpes simplex viruses (HSV) as human pathogens and the emerging prospect of using mutant derivatives of HSV-1 as potential anti-cancer therapeutics have necessitated a thorough investigation into the molecular basis of host-cell permissiveness to HSV. Here we show that NIH-3T3 cells transformed with the oncogenes v-erbB, activated sos or activated ras become significantly more permissive to HSV-1. Inhibitors of the Ras signalling pathway, such as farnesyl transferase inhibitor 1 and PD98059, effectively suppressed HSV-1 infection of ras-transformed cells. Enhanced permissiveness of the transformed cells was linked to the inhibition of virus-induced activation (phosphorylation) of the double-stranded RNA-activated protein kinase (PKR), thereby allowing viral transcripts to be translated in these cells. An HSV-1-derived oncolytic mutant, R3616, was also found to infect preferentially both transformed cells and PKR-/- (but not PKR+/+) mouse embryo fibroblasts. These observations suggest that HSV-1 specifically targets cells with an activated Ras signalling pathway, and have important ramifications in the use of engineered HSV in cancer therapy, the development of strategies against HSV infections, and the controversial role of HSV in human cancers.

Chloride channel ClC-3 promotion of osteogenic differentiation through Runx2.

ClC-3 chloride channel has been speculated to contribute to the acidification of synaptic vesicles and endosomes. However, the biological function of ClC-3 in osteogenesis remains to be determined. In this study, we first analyzed ClC-3 expression in MC3T3-E1 cells and primary mouse osteoblasts and then performed the osteoinductive procedure to determine the effects on gene expression. Subsequently, we transiently transfected ClC-3 cDNA or ClC-3-siRNA into MC3T3-E1 cells to determine the changed phenotype and gene expression. Lastly, we assessed the underlying mechanism responsible for ClC-3-induced osteodifferentiation. We found that ClC-3 mRNA was expressed in primary mouse osteoblasts and MC3T3-E1 cells and induced by using an osteoinductive procedure. We also found that overexpression of ClC-3 contributed to osteodifferentiation, such as increase in the expression of osteogenic markers [alkaline phosphatase (Alp), osteocalcin (Oc), bone sialoprotein (Bsp), osterix (Osx), and runt-related transcription factor 2 (Runx2)], morphological changes, and mineralized nodules in MC3T3-E1 cells. ClC-3 gene silencing suppressed gene expression of these osteogenic markers. Moreover, overexpressed ClC-3 protein co-localized with TGF-beta1 in intracellular organelles, inhibited TGF-beta1 protein expression and induced endosomal acidification. Nevertheless, knockdown of Runx2 expression antagonized the effects of ClC-3 in osteodifferentiation and expression of osteogenic markers. The data from the current study suggest that the function of ClC-3 in osteodifferentiation may be through the Runx2 pathway.

A microfluidic magnetic bead impact generator for physical stimulation of osteoblast cell.

We developed a novel microfluidic cell culture device in which magnetic beads repetitively collide with osteoblast cells, MC3T3-E1, owing to attractive forces generated by pulsed electromagnetic fields and consequently the cells were physically stimulated by bead impacts. Our device consists of an on-chip microelectromagnet and a microfluidic channel which were fabricated by a microelectromechanical system technique. The impact forces and stresses acting on a cell were numerically analyzed and experimentally generated with different sizes of bead (4.5, 7.6 and 8.4 mum) and at various pulse frequencies (60 Hz, 1 kHz and 1 MHz). Cells were synchronized at each specific phase of the cell cycle before stimulation in order to determine the most susceptible phase against bead impacts. The cells were stimulated with different sizes of bead at various pulse frequencies for 1 min at G1, S and G2 phases, respectively, and then counted immediately after one doubling time. The growth rate of cells was highly accelerated when they were stimulated with 4.5 mum beads at G1 phase and a pulse frequency of 1 MHz. Almost all of the cells were viable after stimulation, indicating that our cell stimulator did not cause any cellular damage and is suitable for use in new physical stimulus modalities.

Nonirradiated human fibroblasts and irradiated 3T3-J2 murine fibroblasts as a feeder layer for keratinocyte growth and differentiation in vitro on a fibrin substrate.

The standard method for producing graftable epithelia relies on the presence of a feeder layer of lethally irradiated 3T3-J2 murine fibroblasts (Rheinwald and Green technique). Here, we studied a new keratinocyte culture system, which envisages the utilization of nonirradiated human fibroblasts embedded into a fibrin substrate, in cultures destined for a future clinical application. We tested this culture system using keratinocytes grown on a fibrin gel precoated with 3T3-J2 murine fibroblasts as a control. In order to evaluate the new technology, we compared the clonogenic potential and the proliferative, differentiative and metabolic characteristics of keratinocytes cultured on the fibrin gel under the two culture conditions. The results demonstrated that the proposed technology did not impair the behavior of cultured keratinocytes and revealed that cells maintained their proliferative potential and phenotype under the experimental conditions. In particular, the demonstration of stem cell maintenance under the adopted culture conditions is very important for acute burn treatment with skin substitutes. This work is a first step in the evaluation of a new keratinocyte culture system, which has been studied in order to take advantage of an additional human cell population (i.e. nonirradiated, growing fibroblasts) for future transplantation purposes in acute and chronic wounds. Additional research will allow us to attain (1) the removal of murine cells in the initial phase of keratinocyte cultures, and (2) the removal of other potentially dangerous animal-derived materials from the entire culture system.

PTHrP 1-141 and 1-86 increase in vitro bone formation.

BACKGROUND: Parathyroid hormone-related protein (PTHrP) has anabolic effects in bone, which has led to the clinical use of N-terminal fragments of PTHrP and PTH. Since 10% to 20% of fractures demonstrate healing complications and osteoporosis continues to be a debilitating disease, the development of bone-forming agents is of utmost importance. Due to evidence that regions of PTHrP other than the N-terminus may have bone-forming effects, this study was designed to compare the effects of full-length PTHrP 1-141 to N-terminal PTHrP 1-86 on in vitro bone formation. MATERIALS AND METHODS: MC3T3-E1 pre-osteoblasts were treated once every 6 d for 36 d with 5, 25, and 50 pM of PTHrP 1-141 or 1-86 for 1 or 24 h. Cells were also treated after blocking the N-terminus, the nuclear localization sequence (NLS), and the C-terminus of PTHrP, individually and in combination. Area of mineralization, alkaline phosphatase (ALP), and osteocalcin (OCN) were measured. RESULTS: PTHrP 1-141 and 1-86 increased mineralization after 24-h treatments, but not 1-h. PTHrP 1-141 was more potent than 1-86. Treatment with PTHrP 1-141 for 24-h, but not 1-86, resulted in a concentration-dependent increase in ALP, with no effect after 1-h. Exposure to both peptides for 1- or 24-h induced a concentration-dependent increase in OCN, with 24-h exceeding 1-h. Antibody blocking revealed that the NLS and C-terminus are anabolic. CONCLUSIONS: Both PTHrP 1-141 and 1-86 increased in vitro bone formation; however, PTHrP 1-141 was more effective. The NLS and C-terminus have anabolic effects distinct from the N-terminus. This demonstrates the advantage of PTHrP 1-141 as a skeletal anabolic agent.

Quantification of enhanced osteoblastic adhesion to ultraviolet-treated titanium plate.

Although the advantage of ultraviolet (UV) irradiation of titanium plates for the attachment of osteoblast is known, the details of the experimental conditions have not been described in previous literature. We established optimal conditions of UV irradiation of titanium plate for the adhesion of mouse osteoblast MC3T3-E1 cells. The viable cell number was determined by MTT method. UV irradiation at two different wavelengths (253.7 and 365 nm) enhanced the cell attachment on titanium plate to comparable extents. The optimal UV exposure duration was 20 minutes and prolonged irradiation slightly reduced cell attachment. The attached cells proliferated during 24 hours, accompanied by the enhanced consumption of extracellular glutamine and arginine. The present study supports the previous reports of the efficacy of UV irradiation, and this simple and rapid assay system may be applicable for the study of the interaction of osteoblast and UV-activated titanium plates.

Type of cell death induced by seven metals in cultured mouse osteoblastic cells.

The use of dental metal alloys in the daily clinic makes it necessary to evaluate the cytotoxicity of eluted metal components against oral cells. However, the cytotoxic mechanism and the type of cell death induced by dental metals in osteoblasts have not been well characterized. This study investigated the cytotoxicity of seven metals against the mouse osteoblastic cell line MC3T3-E1. alpha-MEM was used as a culture medium, since this medium provided much superior proliferation of MC3T3-E1 cells over DMEM. Ag (NH(3))(2)F was the most cytotoxic, followed by CuCl>CuCl(2) >CoCl(2), NiCl(2)>FeCl(3) and FeCl(2) (least toxic). None of the metals showed any apparent growth stimulating effect (so-called 'hormesis') at lower concentrations. A time course study demonstrated that two hours of contact between oral cells and Ag (NH(3))(2)F, CuCl, CoCl(2) or NiCl(2) induced irreversible cell death. Contact with these metals induced a smear pattern of DNA fragmentation without activation of caspase-3. Preincubation of MC3T3-E1 cells with either a caspase inhibitor (Z-VAD-FMK) or autophagy inhibitors (3-methyladenine, bafilomycin) failed to rescue them from metal cytotoxicity. These data suggest the induction of necrotic cell death rather than apoptosis and autophagy by metals in this osteoblastic cell line.

Microscale Collagen and Fibroblast Interactions Enhance Primary Human Hepatocyte Functions in Three-Dimensional Models.

Human liver models that are three-dimensional (3D) in architecture are indispensable for compound metabolism/toxicity screening, to model liver diseases for drug discovery, and for cell-based therapies; however, further development of such models is needed to maintain high levels of primary human hepatocyte (PHH) functions for weeks to months. Therefore, here we determined how microscale 3D collagen I presentation and fibroblast interaction affect the longevity of PHHs. High-throughput droplet microfluidics was utilized to generate reproducibly sized (∼300-µm diameter) microtissues containing PHHs encapsulated in collagen I ± supportive fibroblasts, namely, 3T3-J2 murine embryonic fibroblasts or primary human hepatic stellate cells (HSCs); self-assembled spheroids and bulk collagen gels (macrogels) containing PHHs served as controls. Hepatic functions and gene expression were subsequently measured for up to 6 weeks. We found that microtissues placed within multiwell plates rescued PHH functions at 2- to 30-fold higher levels than spheroids or macrogels. Further coating of PHH microtissues with 3T3-J2s led to higher hepatic functions than when the two cell types were either coencapsulated together or when HSCs were used for the coating instead. Importantly, the 3T3-J2-coated PHH microtissues displayed 6+ weeks of relatively stable hepatic gene expression and function at levels similar to freshly thawed PHHs. Lastly, microtissues responded in a clinically relevant manner to drug-mediated cytochrome P450 induction or hepatotoxicity. In conclusion, fibroblast-coated collagen microtissues containing PHHs display high hepatic functions for 6+ weeks and are useful for assessing drug-mediated CYP induction and hepatotoxicity. Ultimately, microtissues may find utility for modeling liver diseases and as building blocks for cell-based therapies.

Insulin causes fatty acid transport protein translocation and enhanced fatty acid uptake in adipocytes.

Fatty acid uptake into 3T3 L1 adipocytes is predominantly transporter mediated. Here we show that, during 3T3 L1 adipocyte differentiation, expression of fatty acid transport proteins (FATPs) 1 and 4 is induced. Using subcellular membrane fractionation and immunofluorescence microscopy, we demonstrate that, in adipocytes, insulin induces plasma membrane translocation of FATPs from an intracellular perinuclear compartment to the plasma membrane. This translocation was observed within minutes of insulin treatment and was paralleled by an increase in long chain fatty acid (LCFA) uptake. In contrast, treatment with TNF-alpha inhibited basal and insulin-induced LCFA uptake and reduced FATP1 and -4 levels. Thus, hormonal regulation of FATP activity may play an important role in energy homeostasis and metabolic disorders such as type 2 diabetes.

Conjugated linoleic acid activates AMP-activated protein kinase and reduces adiposity more effectively when used with metformin in mice.

Trans-10, cis-12 (t10c12) conjugated linoleic acid (CLA) reduces lipid levels in adipocytes, but the mechanisms involved are still emerging. The hypotheses of this study were that t10c12 CLA treatment activated AMP-activated protein kinase (AMPK) and that the effectiveness of a low dose of t10c12 CLA would be increased when combined with an AMPK activator. We demonstrated t10c12 CLA, directly or indirectly, activated AMPK and increased the amount of phosphorylated acetyl-CoA carboxylase (ACC) in 3T3-L1 adipocytes. Compound C, a potent inhibitor of AMPK, attenuated the phosphorylation of ACC, integrated stress response (ISR), inflammatory response, reduction in key lipogenic transcription factors, and triglyceride (TG) reduction that otherwise occurred in t10c12 CLA-treated adipocytes. Treatment of adipocytes or mice with a low dose of t10c12 CLA in conjunction with the AMPK activator metformin resulted in more TG loss than treatment with the individual chemicals. Additionally, although an inflammatory response was required for robust TG reduction, the combination of t10c12 CLA with AMPK activators had a similar TG loss with a reduced inflammatory response. A microarray analysis of the transcriptional response to either t10c12 CLA, metformin, or the combination, indicated the responses were very similar, with a correlation coefficient of 0.91 or better for genes in the ISR or lipid-related pathways. Altogether, these results support our hypotheses that t10c12 CLA activates AMPK, directly or indirectly, and that metformin increases the effectiveness of t10c12 CLA in reducing TG amounts in adipocytes.

Cell surface expression of receptor protein tyrosine phosphatase RPTP mu is regulated by cell-cell contact.

RPTP mu is a transmembrane protein tyrosine phosphatase with an adhesion molecule-like ectodomain. It has recently been shown that RPTP mu mediates homophilic interactions when expressed in insect cells. In this study, we have examined how RPTP mu may function as a cell contact receptor in mink lung epithelial cells, which express RPTPmu endogenously, as well as in transfected 3T3 cells. We find that RPTP mu has a relatively short half-life (3-4 hours) and undergoes posttranslational cleavage into two noncovalently associated subunits, with both cleaved and uncleaved molecules being present on the cell surface (roughly at a 1:1 ratio); shedding of the ectodomain subunit is observed in exponentially growing cells. Immunofluorescence analysis reveals that surface expression of RPTPmu is restricted to regions of tight cell-cell contact. RPTPmu surface expression increases significantly with increasing cell density. This density-induced upregulation of RPTP mu is independent of its catalytic activity and is also observed when transcription is driven by a constitutive promoter, indicating that modulation of RPTPmu surface expression occurs posttranscriptionally. Based on our results, we propose the following model of RPTP mu function: In the absence of cell-cell contact, newly synthesized RPTP mu molecules are rapidly cleared from the cell surface. Cell-cell contact causes RPTPmu to be trapped at the surface through homophilic binding, resulting in accumulation of RPTP mu at intercellular contact regions. This contact-induced clustering of RPTPmu may then lead to tyrosine dephosphorylation of intracellular substrates at cell-cell contacts.

Molecular regulation of GLUT-4 targeting in 3T3-L1 adipocytes.

Insulin stimulates glucose transport in muscle and adipose tissue by triggering the movement of the glucose transporter GLUT-4 from an intracellular compartment to the cell surface. Fundamental to this process is the intracellular sequestration of GLUT-4 in nonstimulated cells. Two distinct targeting motifs in the amino and carboxy termini of GLUT-4 have been previously identified by expressing chimeras comprised of portions of GLUT-4 and GLUT-1, a transporter isoform that is constitutively targeted to the cell surface, in heterologous cells. These motifs-FQQI in the NH2 terminus and LL in the COOH terminus-resemble endocytic signals that have been described in other proteins. In the present study we have investigated the roles of these motifs in GLUT-4 targeting in insulin-sensitive cells. Epitope-tagged GLUT-4 constructs engineered to differentiate between endogenous and transfected GLUT-4 were stably expressed in 3T3-L1 adipocytes. Targeting was assessed in cells incubated in the presence or absence of insulin by subcellular fractionation. The targeting of epitope-tagged GLUT-4 was indistinguishable from endogenous GLUT-4. Mutation of the FQQI motif (F5 to A5) caused GLUT-4 to constitutively accumulate at the cell surface regardless of expression level. Mutation of the dileucine motif (L489L490 to A489A490) caused an increase in cell surface distribution only at higher levels of expression, but the overall cells surface distribution of this mutant was less than that of the amino-terminal mutants. Both NH2- and COOH-terminal mutants retained insulin-dependent movement from an intracellular to a cell surface locale, suggesting that neither of these motifs is involved in the insulin-dependent redistribution of GLUT-4. We conclude that the phenylalanine-based NH2-terminal and the dileucine-based COOH-terminal motifs play important and distinct roles in GLUT-4 targeting in 3T3-L1 adipocytes.

Suppression of vinculin expression by antisense transfection confers changes in cell morphology, motility, and anchorage-dependent growth of 3T3 cells.

The expression of vinculin, a major component of adhesion plaques and cell-cell junctions, is markedly modulated in cells during growth activation, differentiation, motility and cell transformation. The stimulation of quiescent cells by serum factors and the culturing of cells on highly adhesive matrices induce vinculin gene expression, whereas the transformation of fibroblast and epithelial cells often results in decreased vinculin expression (reviewed in Rodríguez Fernández, J. L., B. Geiger, D. Salomon, I. Sabanay, M. Zöller, and A. Ben-Ze'ev. 1992. J. Cell Biol. 119:427). To study the effect of reduced vinculin expression on cell behavior, 3T3 cells were transfected with an antisense vinculin cDNA construct, and clones displaying decreased vinculin levels down to 10-30% of control levels were isolated. These cells showed a round phenotype with smaller and fewer vinculin-positive plaques localized mostly at the cell periphery. In addition, they displayed an increased motility compared to controls, manifested by a faster closure of "wounds" introduced into the monolayer, and by the formation of longer phagokinetic tracks. Moreover, the antisense transfectants acquired a higher cloning efficiency and produced larger colonies in soft agar than the parental counterparts. The results demonstrate that the regulation of vinculin expression in cells can affect, in a major way, cell shape and motility, and that decreased vinculin expression can induce cellular changes reminiscent of those found in transformed cells.

The assembly of integrin adhesion complexes requires both extracellular matrix and intracellular rho/rac GTPases.

Interaction of cells with extracellular matrix via integrin adhesion receptors plays an important role in a wide range of cellular: functions, for example cell growth, movement, and differentiation. Upon interaction with substrate, integrins cluster and associate with a variety of cytoplasmic proteins to form focal complexes and with the actin cytoskeleton. Although the intracellular signals induced by integrins are at present undefined, it is thought that they are mediated by proteins recruited to the focal complexes. It has been suggested, for example, that after recruitment to focal adhesions p125FAK can activate the ERK1/2 MAP kinase cascade. We have previously reported that members of the rho family of small GTPases can trigger the assembly of focal complexes when activated in cells. Using microinjection techniques, we have now examined the role of the extracellular matrix and of the two GTP-binding proteins, rac and rho, in the assembly of integrin complexes in both mouse and human fibroblasts. We find that the interaction of integrins with extracellular matrix alone is not sufficient to induce integrin clustering and focal complex formation. Similarly, activation of rho or rac by extracellular growth factors does not lead to focal complex formation in the absence of matrix. Focal complexes are only assembled in the presence of both matrix and functionally active members of the rho family. In agreement with this, the interaction of integrins with matrix in the absence of rho/rac activity is unable to activate the ERK1/2 kinases in Swiss 3T3 cells. In fact, ERK1/2 can be activated fully by growth factors in the absence of matrix and it seems unlikely, therefore, that the adhesion dependence of fibroblast growth is mediated through the ras/MAP kinase pathway. We conclude that extracellular matrix is not sufficient to trigger focal complex assembly and subsequent integrin-dependent signal transduction in the absence of functionally active members of the rho family of GTPases.