N-cadherin/FGFR promotes metastasis through epithelial-to-mesenchymal transition and stem/progenitor cell-like properties.

N-cadherin and HER2/neu were found to be co-expressed in invasive breast carcinomas. To test the contribution of N-cadherin and HER2 in mammary tumor metastasis, we targeted N-cadherin expression in the mammary epithelium of the MMTV-Neu mouse. In the context of ErbB2/Neu, N-cadherin stimulated carcinoma cell invasion, proliferation and metastasis. N-cadherin caused fibroblast growth factor receptor (FGFR) upmodulation, resulting in epithelial-to-mesenchymal transition (EMT) and stem/progenitor like properties, involving Snail and Slug upregulation, mammosphere formation and aldehyde dehydrogenase activity. N-cadherin potentiation of the FGFR stimulated extracellular signal regulated kinase (ERK) and protein kinase B (AKT) phosphorylation resulting in differential effects on metastasis. Although ERK inhibition suppressed cyclin D1 expression, cell proliferation and stem/progenitor cell properties, it did not affect invasion or EMT. Conversely, AKT inhibition suppressed invasion through Akt 2 attenuation, and EMT through Snail inhibition, but had no effect on cyclin D1 expression, cell proliferation or mammosphere formation. These findings suggest N-cadherin/FGFR has a pivotal role in promoting metastasis through differential regulation of ERK and AKT, and underscore the potential for targeting the FGFR in advanced ErbB2-amplified breast tumors.

N-cadherin regulates mammary tumor cell migration through Akt3 suppression.

N-cadherin is a cell-cell adhesion molecule that plays a role in breast cancer metastasis. Here, we show that in vivo expression of N-cadherin in the PyMT mouse model, which enhances mammary tumor metastasis, results in selective inhibition of Akt3 expression and phosphorylation. Similarly, exogenous expression of N-cadherin in PyMT or MCF-7 mammary tumor cells enhanced cell motility and caused a dramatic reduction in Akt3 expression and phosphorylation. Moreover, knockdown of Akt3 in PyMT tumor cells increased cell motility and disrupted mammary morphogenesis, but had no effect on cell proliferation. Conversely, overexpression of wild-type Akt3 in PyMT-N-cadherin cells inhibited cell motility promoted by N-cadherin. Taken altogether, these findings demonstrate that N-cadherin suppresses Akt3 to promote cell motility and highlight the intricate regulation of Akt isoforms by N-cadherin during metastasis.

p21CIP1 mediates reciprocal switching between proliferation and invasion during metastasis.

Cell proliferation and invasion are critical for malignant progression, yet how these processes relate to each other and whether they regulate one another during metastasis is unknown. We show that invasiveness of breast cancer cells is associated with growth arrest due to p21CIP1 upregulation. Knockdown of p21CIP1 increases cell proliferation and suppresses invasion. Since p21CIP1 acts to inhibit cyclin E during cell-cycle progression, we demonstrated that a constitutively active form of cyclin E had similar effects to p21CIP1 inhibition resulting in enhanced cell growth and suppressed invasiveness. We tested these findings in vivo in the Polyoma middle T mammary tumor model in which p21CIP1 was deleted. p21CIP1 knockout mice exhibited dramatic suppression of metastasis, independent of tumor growth, which was rescued by p21CIP1. Metastasis suppression by p21CIP1 ablation was associated with striking cytoskeletal reorganization leading to a non-invasive and highly proliferative state. Thus, p21CIP1 regulates metastasis by mediating reciprocal switching between invasion and proliferation.

Quorum sensing differentially regulates Pseudomonas aeruginosa type VI secretion locus I and homologous loci II and III, which are required for pathogenesis.

Pseudomonas aeruginosa harbours three type VI secretion (T6S) loci. Although HSI-I has been partially studied, limited knowledge is available on the homologous loci HSI-II and HSI-III. We show that quorum sensing (QS) differentially regulates the expression of genes at all three loci. HSI-I-associated gene expression is suppressed by both the homoserine lactone transcription factor LasR and the 4-hydroxy-2-alkylquinoline (HAQ) transcriptional regulator MvfR. Conversely, both HSI-II and HSI-III loci are positively controlled by LasR and MvfR. PqsE, a key component of the MvfR regulon, is required for the expression of part of HSI-III but not HSI-II, and previously identified inhibitors of HAQ biosynthesis significantly downregulate HSI-II and -III gene expression. Animal and plant infection studies reveal that both HSI-II and -III play important roles in pathogenesis. Furthermore, analysis of a double DeltaHSI-II : : III mutant suggests that these loci functionally compensate for one another in virulence. This study illustrates the contribution of the QS systems to T6S gene regulation and reveals the importance of HSI-II and -III in mediating P. aeruginosa pathogenesis. Moreover, this work provides new insights into the design and development of selective compounds that may restrict human P. aeruginosa and possibly other clinical infections.

Escherichia coli mazEF-mediated cell death as a defense mechanism that inhibits the spread of phage P1.

The Escherichia coli gene pair mazEF is a regulatable chromosomal toxin-antitoxin module: mazF encodes a stable toxin and mazE encodes for a labile antitoxin that overcomes the lethal effect of MazF. Because MazE is labile, inhibition of mazE expression results in cell death. We studied the effect of mazEF on the development of bacteriophage P1 upon thermoinduction of the prophage P1CM c1ts and upon infection with virulent phage particles (P1vir). In several E. coli strains, we showed that the Delta mazEF derivative strains produced significantly more phages than did the parent strain. In addition, upon induction of K38(P1CM c1ts), nearly all of the Delta mazEF mutant cells lysed; in contrast, very few of the parental mazEF + K38 cells underwent lysis. However, most of these cells did not remain viable. Thus, while the Delta mazEF cells die as a result of the lytic action of the phage, most of the mazEF+ cells are killed by a different mechanism, apparently through the action of the chromosomal mazEF system itself. Furthermore, the introduction of lysogens into a growing non-lysogenic culture is lethal to Delta mazEF but not for mazEF+ cultures. Thus, although mazEF action causes individual cells to die, upon phage growth this is generally beneficial to the bacterial culture because it causes P1 phage exclusion from the bacterial population. These results provide additional support for the view that bacterial cultures may share some of the characteristics of multicellular organisms.

Programmed cell death in Escherichia coli: some antibiotics can trigger mazEF lethality.

The discovery of toxin-antitoxin gene pairs (also called addiction modules) on extrachromosomal elements of Escherichia coli, and particularly the discovery of homologous modules on the bacterial chromosome, suggest that a potential for programmed cell death may be inherent in bacterial cultures. We have reported on the E. coli mazEF system, a regulatable addiction module located on the bacterial chromosome. MazF is a stable toxin and MazE is a labile antitoxin. Here we show that cell death mediated by the E. coli mazEF module can be triggered by several antibiotics (rifampicin, chloramphenicol, and spectinomycin) that are general inhibitors of transcription and/or translation. These antibiotics inhibit the continuous expression of the labile antitoxin MazE, and as a result, the stable toxin MazF causes cell death. Our results have implications for the possible mode(s) of action of this group of antibiotics.

Postsegregational killing mediated by the P1 phage "addiction module" phd-doc requires the Escherichia coli programmed cell death system mazEF.

"Addiction modules" consist of two genes; the product of the second is long lived and toxic, while the product of the first is short lived and antagonizes the lethal action of the toxin. The extrachromosomal addiction module phd-doc, located on the P1 prophage, is responsible for the postsegregational killing effect (death of plasmid-free cells). The Escherichia coli chromosomal addiction module analogue, mazEF, is responsible for the induction of programmed cell death. Here we show that the postsegregational killing mediated by the P1 phd-doc module depends on the presence of the E. coli mazEF system. In addition, we demonstrate that under conditions of postsegregational killing, mediated by phd-doc, protein synthesis of E. coli is inhibited. Based on our findings, we suggest the existence of a coupling between the phd-doc and mazEF systems.

Basement membrane alterations in psoriasis are accompanied by epidermal overexpression of MMP-2 and its inhibitor TIMP-2.

Psoriasis is most probably an inherited disease characterized by cell proliferation, angiogenesis, and an inflammatory process. The pathophysiology remains unknown, although an alteration in cell-cell and cell-matrix adhesion versus an autoimmune process has been proposed as the primary defect. Here, we show evidence of a new mechanism involving basement membrane alterations accompanied by keratinocyte overexpression of matrix metalloproteinase (MMP) 2 and tissue inhibitor of MMP-2 (TIMP-2) in both uninvolved and involved psoriatic skin. Immunocytochemistry with antibodies against collagen IV (alpha1, alpha2 chains) and laminins (alpha2, alpha5, beta1, gamma1 chains) revealed gaps, folding, and reduplication of the epidermo-dermal basement membrane. There was overexpression of MMP-2 in the cytoplasm of suprabasal keratinocytes. Gelatin zymography revealed pro-MMP-2 and its activated form, a-MMP-2, in both uninvolved and involved psoriatic skin, whereas pro-MMP-9 was only present in involved skin. TIMP-2 was expressed at the cell surface of psoriatic involved suprabasal keratinocytes whereas it was restricted to basal keratinocytes in uninvolved areas. Western blots showed a marked increase in a-MMP-2 and TIMP-2 in uninvolved and involved psoriatic skin although it was more pronounced in the latter. MT1-MP, known to activate pro-MMP-2, was increased in involved areas. In situ hybridization revealed strong signals of MMP-2 mRNA in both uninvolved and involved psoriatic epidermis. The overexpression of MMP-2 in uninvolved and involved psoriatic epidermis supports the concept that the primary alteration may reside in the keratinocyte. In addition, the presence of the activated form of MMP-2 could be responsible for cell-cell and cell-matrix changes noted in psoriatic epidermis.

Exogenous expression of N-cadherin in breast cancer cells induces cell migration, invasion, and metastasis.

E- and N-cadherin are calcium-dependent cell adhesion molecules that mediate cell-cell adhesion and also modulate cell migration and tumor invasiveness. The loss of E-cadherin-mediated adhesion has been shown to play an important role in the transition of epithelial tumors from a benign to an invasive state. However, recent evidence indicates that another member of the cadherin family, N-cadherin, is expressed in highly invasive tumor cell lines that lacked E-cadherin expression. These findings have raised the possibility that N-cadherin contributes to the invasive phenotype. To determine whether N-cadherin promotes invasion and metastasis, we transfected a weakly metastatic and E-cadherin-expressing breast cancer cell line, MCF-7, with N-cadherin and analyzed the effects on cell migration, invasion, and metastasis. Transfected cells expressed both E- and N-cadherin and exhibited homotypic cell adhesion from both molecules. In vitro, N-cadherin-expressing cells migrated more efficiently, showed an increased invasion of Matrigel, and adhered more efficiently to monolayers of endothelial cells. All cells produced low levels of the matrix metalloproteinase MMP-9, which was dramatically upregulated by treatment with FGF-2 only in N-cadherin-expressing cells. Migration and invasion of Matrigel were also greatly enhanced by this treatment. When injected into the mammary fat pad of nude mice, N-cadherin-expressing cells, but not control MCF-7 cells, metastasized widely to the liver, pancreas, salivary gland, omentum, lung, lymph nodes, and lumbar spinal muscle. The expression of both E- and N-cadherin was maintained both in the primary tumors and metastatic lesions. These results demonstrate that N-cadherin promotes motility, invasion, and metastasis even in the presence of the normally suppressive E-cadherin. The increase in MMP-9 production by N-cadherin-expressing cells in response to a growth factor may endow them with a greater ability to penetrate matrix protein barriers, while the increase in their adherence to endothelium may improve their ability to enter and exit the vasculature, two properties that may be responsible for metastasis of N-cadherin-expressing cells.

The epidermal growth factor receptor modulates the interaction of E-cadherin with the actin cytoskeleton.

Alterations in the expression or function of molecules that affect cellular adhesion and proliferation are thought to be critical events for tumor progression. Loss of expression of the cell adhesion molecule E-cadherin and increased expression of the epidermal growth factor receptor are two prominent molecular events that are associated with tumorigenesis. The regulation of E-cadherin-dependent cell adhesion by epidermal growth factor (EGF) was therefore examined in the human breast cancer cell line, MDA-MB-468. In this study, changes were observed in the subcellular distribution of components that mediate the cytoplasmic connection between E-cadherin and the actin-based cytoskeleton in response to activation of the EGF receptor. Serum withdrawal activated E-cadherin-dependent cell-cell aggregation in MDA-MB-468 cells, and this treatment stimulated the interaction of actin, alpha-actinin, and vinculin with E-cadherin complexes, despite the absence of alpha-catenin in these cells. By contrast, the co-precipitation of actin with E-cadherin was not detected in several alpha-catenin positive epithelial cell lines. Treatment with EGF inhibited cellular aggregation but did not affect either the levels of E-cadherin or catenin expression nor the association of catenins (beta-catenin, plakoglobin/gamma-catenin, or p120(cas)) with E-cadherin. However, EGF treatment of the MDA-MB-468 cell line dissociated actin, alpha-actinin, and vinculin from the E-cadherin-catenin complex, and this coincided with a robust phosphorylation of beta-catenin, plakoglobin/gamma-catenin, and p120(cas) on tyrosine residues. Furthermore, inactivation of the EGF receptor in serum-treated MDA-MB-468 cells with either a function-blocking antibody or EGF receptor kinase inhibitors mimicked the effects of serum starvation by stimulating both cellular aggregation and assembly of E-cadherin complexes with vinculin and actin. These results demonstrate that the EGF receptor directly regulates cell-cell adhesion through modulation of the interaction of E-cadherin with the actin cytoskeleton and thus substantiates the coordinate role of both of these molecules in tumor progression and metastasis.

N-cadherin promotes adhesion between invasive breast cancer cells and the stroma.

Calcium-dependent cell adhesion molecules (cadherins) are involved in maintaining the epithelial structure of a number of tissues including the mammary gland. In breast and other tumor types, loss of E-cadherin expression has been seen in high grade tumors and correlates with increased invasiveness. Here we show high levels of expression of N-cadherin in the most invasive breast cancer cell lines which was inversely correlated with their expression of E-cadherin. A stromal cell line also expressed N-cadherin in accordance with its fibroblastic morphology. N-cadherin localized to areas of cell-cell contact in all cells that expressed it. Calcium-dependent intercellular adhesion of N-cadherin-expressing breast cancer and stromal cells was specifically inhibited by an anti N-cadherin monoclonal antibody. In addition, N-cadherin promoted the interaction of invasive breast cancer cells with mammary stromal cells; in contrast, E-cadherin expressing cell lines did not co-aggregate with stromal cells. The combined results suggest a functional role for N-cadherin in cohesion of breast tumor cells which, in addition promotes their interaction with the surrounding stromal cells, thereby facilitating invasion and metastasis.

Vinculin is associated with the E-cadherin adhesion complex.

Cadherins mediate calcium-dependent cell-cell adhesion, and this activity is regulated by cytoplasmic interactions between cadherins, catenins, and the actin-based cytoskeleton. alpha-Catenin plays a critical role in the transmembrane anchorage of cadherins, and deletion of alpha-catenin has been shown to inactivate cadherin-mediated adhesion, resulting in a nonadhesive phenotype. Here we show that serum starvation increases E-cadherin expression and induces E-cadherin-dependent adhesion in the MDA-MB-468 breast cancer cell line. This adhesion occurred despite a lack of alpha-catenin expression, which was caused by mutations in the alpha-catenin gene. Coprecipitation analysis suggests that this adhesion may be mediated by cytoplasmic connections from cadherins to the cytoskeleton involving vinculin. A high level of vinculin associated with E-cadherin immunoprecipitates was observed in MDA-MB-468 cells. In contrast, vinculin was not detected in E-cadherin complexes in the A431 and MCF-7 epithelial carcinoma cell lines, which express alpha-catenin. However, in reciprocal immunoprecipitations using anti-vinculin antibodies, E-cadherin associated strongly with vinculin in MDA-MB-468 cells and, to a lesser extent, in A431 and MCF-7 cells. These results suggest that both alpha-catenin and vinculin may be present in the adhesion complex. To test the hypothesis that vinculin associates with E-cadherin complexes via beta-catenin, excess recombinant beta-catenin or alpha-catenin fusion protein was added to MDA-MB-468 cell lysates. Both specifically inhibited the coprecipitation of E-cadherin with vinculin, suggesting competition for the same binding site. These results suggest that vinculin plays a role in the establishment or regulation of the cadherin-based cell adhesion complex by direct interaction with beta-catenin.

Functional analysis of posttranslational cleavage products of the neuron-glia cell adhesion molecule, Ng-CAM.

Neuron-glia cell adhesion molecule (Ng-CAM) mediates cell adhesion between neurons homophilically and between neurons and glia heterophilically; it also promotes neurite outgrowth. In the chick brain, Ng-CAM is detected as glycoproteins of 190 and 210 kD (Ng-CAM200) with posttranslational cleavage products of 135 kD (F135, which contains most of the extracellular region) and 80 kD (F80, which includes the transmembrane and the cytoplasmic domains). To examine the functions of each of these components, we have expressed Ng-CAM200, F135, and F80 in murine L cells, and F135 and F80 as GST fusion proteins in the pGEX vector in bacteria. Appropriately transfected L cells expressed each of these proteins on their surfaces; F135 was also found in the media of cells transfected with Ng-CAM200 and F135. In addition to binding homophilically, cells transfected with Ng-CAM200 and F135 bound heterophilically to untransfected L cells, suggesting that there is a ligand for Ng-CAM on fibroblasts that may be related to the glial ligand. Detailed studies using the transfected cells and the fusion proteins indicated that both the homophilic and the heterophilic binding activities of Ng-CAM are localized in the F135 fragment of the molecule. The results also indicated that proteolytic cleavage of Ng-CAM200 is not required either for its expression on the cell surface or for cell adhesion and that there is an "anchor" for F135 on L cells (and presumably on neurons). In contrast to the cell binding results, the F80 but not the F135 fusion protein enhanced the outgrowth of neurites from dorsal root ganglion cells; this activity was associated with the FnIII repeats of F80. The observations that a protein corresponding to F135 contains the cell aggregation sites whereas one corresponding to the F80 has the ability to promote neurite outgrowth suggest that proteolytic cleavage may be an important event in regulating these Ng-CAM activities during embryonic development and neural regeneration.

EGF receptor-mediated signals are differentially modulated by concanavalin A.

NIH 3T3 cells expressing high levels of the human epidermal growth factor (EGF) receptor were used to examine the effects of the lectin concanavalin A (Con A) on EGF-mediated signaling events. Proliferation of NIH 3T3 cells expressing high levels of the human EGF receptor was inhibited in a dose-dependent manner by Con A. At the same time, Con A also inhibited both dimerization and tyrosine phosphorylation of the EGF receptor. Tyrosine phosphorylation of the enzyme phospholipase C-gamma, a substrate of the phosphorylated EGF receptor kinase, was also inhibited. In contrast, EGF-stimulated changes in pH, calcium, and levels of inositol phosphates were unaffected by the presence of Con A. These results indicate that certain signals (changes in the levels of intracellular calcium, pH, and inositol phosphates) mediated by EGF binding to its receptor still occur when receptor dimerization and phosphorylation are dramatically decreased, suggesting that multiple independent signals are transmitted by the binding of EGF to its receptor.

Enhancement of bone growth into metal screws implanted in the medullary canal of the femur in rats.

The kinetics of growth of bone into control (nontreated) and heat-treated screws made of stainless steel (type 316L) and Ti-6Al-4V, implanted in the medullary canal of the femur in rats, were studied by mechanical, histological, and biochemical methods. A progressive and significant increase in the ingrowth of bone, as reflected by interfacial shear strengths of the screws, was measured with time after implantation. At all time intervals for as long as 35 days after implantation, the shear strength of the heat-treated Ti-6Al-4V and stainless-steel screws was significantly higher than (1.6-3.4 times) that of the control screws. The specific activity of alkaline phosphatase in extracts of tissue from around the implanted screws peaked 6 days after insertion, with significantly higher values at 5, 6, and 7 days postoperatively for the heat-treated screws than for the controls. The extent of calcification also was higher at all time intervals for the heat-treated screws than for the controls. The histological evaluation of formation of bone between the ridges of the implanted screw corroborated the mechanical and biochemical measurements. At each time interval, a more mature bone was noted around the heat-treated screws than around the controls. It was concluded that the heat treatment of metal implants before insertion can result in augmentation of osseous ingrowth 1.6-5.3 times that into control implants in an in vivo experimental model.

Bone growth to metal implants is regulated by their surface chemical properties.

Bony ingrowth to control (non-treated) and heat-treated stainless steel and Ti-6Al-4V implants into the medullary canal of the femur in rats was studied by mechanical, chemical and Auger electron spectroscopic methods. At all time intervals up to 35 d post-implantation, the shear strengths of the heat-treated Ti-6Al-4V and stainless steel implants were significantly higher (1.6-fold to 3.4-fold) than in control implants. Using Auger electron spectroscopy depth profiling methods, it was found that the heat treatment modified the implant surface composition significantly, resulting in a thicker oxide layer and other chemical changes. It is concluded that heat treatment of metal implants prior to their insertion alters their chemical surface properties and augments bony ingrowth to them.

Epidermal growth factor (EGF) stimulates association and kinase activity of Raf-1 with the EGF receptor.

Raf-1 serine- and threonine-specific protein kinase is transiently activated in cells expressing the epidermal growth factor (EGF) receptor upon treatment with EGF. The stimulated EGF receptor coimmunoprecipitates with Raf-1 kinase and mediates protein kinase C-independent phosphorylation of Raf-1 on serine residues. Hyperphosphorylated Raf-1 has lower mobility on sodium dodecyl sulfate gels and has sixfold-increased activity in immunocomplex kinase assay with histone H1 or Raf-1 sequence-derived peptide as a substrate. Raf-1 activation requires kinase-active EGF receptor; a point mutant lacking tyrosine kinase activity in inactive in Raf-1 coupling and association. It is noteworthy that tyrosine phosphorylation of c-Raf-1 induced by EGF was not detected in these cells. These observations suggest that Raf-1 kinase may act as an important downstream effector of EGF signal transduction.

Immunohistochemical expression of c-erbB-2 in human breast cancer by monoclonal antibody: correlation with lymph node and ER status.

c-erbB-2 Protein expression was investigated in a series of fifty primary breast cancers by means of a specific monoclonal antibody and immunocytochemistry. Specific staining was observed at the plasma membrane level of neoplastic cells, according to the reported localization of c-erbB-2 protein. Sixty-four percent of tumors scored positive, with a variable amount of stained cells. The rate of protein expression was found to exceed the reported gene amplification. No relationship was observed between c-erbB-2 protein staining and age, menopausal status or histologic subtypes. An inverse association was found between c'erbB-2 protein staining and estrogen receptor content of tumors, assayed by immunocytochemistry. A positive relationship was observed between c-erbB-2 protein expression and presence of axillary node metastasis. These findings suggest that c-erbB-2 protein expression is a marker of tumor aggressiveness and that its prognostic power deserves further investigation both in node-positive and node-negative patients.

Identification of autophosphorylation sites of HER2/neu.

HER2 or c-erbB-2 is a putative growth factor receptor with sequence homology to the epidermal growth factor receptor. It is the human homologue of the rat protooncogene neu and may have an important role in human malignancies such as breast and ovarian cancers. Like other growth factor receptors, HER2 has intrinsic protein tyrosine kinase activity and undergoes autophosphorylation. Recently, we have demonstrated that, similar to the epidermal growth factor receptor, all autophosphorylation sites of HER2 are localized in the carboxyl terminus of this protein. In the present study, immunopurified HER2 was allowed to autophosphorylate, and tryptic phosphopeptides were generated. After purification of these phosphopeptides by high performance liquid chromatography, microsequencing was performed. Utilizing this approach, two autophosphorylation sites were unequivocally identified at Y1023 and Y1248. The sequences of two other tyrosine phosphorylated tryptic peptides were determined, but the exact site of autophosphorylation could not be determined because multiple tyrosines were located on each peptide. However, each of these peptides contains tyrosines that correspond to major autophosphorylation sites of the epidermal growth factor receptor, suggesting that, in addition to Y1023 and Y1248, Y1139 and Y1222 also serve as autophosphorylation sites of HER2.

Pathologic findings from the National Surgical Adjuvant Breast and Bowel Project: prognostic significance of erbB-2 protein overexpression in primary breast cancer.

In order to investigate the prognostic significance of erbB-2 overexpression, immunohistochemical staining for the erbB-2 protein was performed on sections from paraffin blocks of 292 primary invasive breast cancers obtained from women enrolled in the National Surgical Adjuvant Breast and Bowel Project (NSABP) protocol B-06. Positive reaction indicative of erbB-2 overexpression was observed on tumor cells in 62 (21%) samples. Women whose cancers were judged to have erbB-2 overexpression had a significantly worse overall survival (P = .0012) with twice the mortality rate of women without detectable erbB-2 expression. No statistically significant effect was evident for disease-free survival (P = .22). In multivariate analysis, detection of erbB-2 overexpression was the second most predictive independent variable for survival after nodal status. Overexpression of erbB-2 was more common among tumors of poor nuclear grade (29%) than those of good nuclear grade (12%). The association of erbB-2 overexpression with decreased survival was evident only among women with tumors of good nuclear grade. In this subgroup, erbB-2 overexpression was associated with an approximately fivefold increase in mortality rate (P = .00001). The combined predictive value of erbB-2 overexpression and nuclear grade was evident regardless of their lymph node status. These results provide evidence that detection of erbB-2 overexpression may be an independent prognostic variable for patient survival. Moreover, when combined with evaluation of nuclear grade, it may be possible to use immunostaining for erbB-2 protein to identify patients at increased risk from within a relatively low-risk group.