High-Speed Time-Domain Diffuse Optical Tomography with a Sensitivity Equation-based Neural Network.

Steady progress in time-domain diffuse optical tomography (TD-DOT) technology is allowing for the first time the design of low-cost, compact, and high-performance systems, thus promising more widespread clinical TD-DOT use, such as for recording brain tissue hemodynamics. TD-DOT is known to provide more accurate values of optical properties and physiological parameters compared to its frequency-domain or steady-state counterparts. However, achieving high temporal resolution is still difficult, as solving the inverse problem is computationally demanding, leading to relatively long reconstruction times. The runtime is further compromised by processes that involve 'nontrivial' empirical tuning of reconstruction parameters, which increases complexity and inefficiency. To address these challenges, we present a new reconstruction algorithm that combines a deep-learning approach with our previously introduced sensitivity-equation-based, non-iterative sparse optical reconstruction (SENSOR) code. The new algorithm (called SENSOR-NET) unfolds the iterations of SENSOR into a deep neural network. In this way, we achieve high-resolution sparse reconstruction using only learned parameters, thus eliminating the need to tune parameters prior to reconstruction empirically. Furthermore, once trained, the reconstruction time is not dependent on the number of sources or wavelengths used. We validate our method with numerical and experimental data and show that accurate reconstructions with 1 mm spatial resolution can be obtained in under 20 milliseconds regardless of the number of sources used in the setup. This opens the door for real-time brain monitoring and other high-speed DOT applications.

Unrolled-DOT: an interpretable deep network for diffuse optical tomography.

Significance: Imaging through scattering media is critical in many biomedical imaging applications, such as breast tumor detection and functional neuroimaging. Time-of-flight diffuse optical tomography (ToF-DOT) is one of the most promising methods for high-resolution imaging through scattering media. ToF-DOT and many traditional DOT methods require an image reconstruction algorithm. Unfortunately, this algorithm often requires long computational runtimes and may produce lower quality reconstructions in the presence of model mismatch or improper hyperparameter tuning. Aim: We used a data-driven unrolled network as our ToF-DOT inverse solver. The unrolled network is faster than traditional inverse solvers and achieves higher reconstruction quality by accounting for model mismatch. Approach: Our model "Unrolled-DOT" uses the learned iterative shrinkage thresholding algorithm. In addition, we incorporate a refinement U-Net and Visual Geometry Group (VGG) perceptual loss to further increase the reconstruction quality. We trained and tested our model on simulated and real-world data and benchmarked against physics-based and learning-based inverse solvers. Results: In experiments on real-world data, Unrolled-DOT outperformed learning-based algorithms and achieved over 10× reduction in runtime and mean-squared error, compared to traditional physics-based solvers. Conclusion: We demonstrated a learning-based ToF-DOT inverse solver that achieves state-of-the-art performance in speed and reconstruction quality, which can aid in future applications for noninvasive biomedical imaging.

Optical scattering as an early marker of apoptosis during chemotherapy and antiangiogenic therapy in murine models of prostate and breast cancer.

Monitoring of the in vivo tumor state to track therapeutic response in real time may help to evaluate new drug candidates, maximize treatment efficacy, and reduce the burden of overtreatment. Current preclinical tumor imaging methods have largely focused on anatomic imaging (e.g., MRI, ultrasound), functional imaging (e.g., FDG-PET), and molecular imaging with exogenous contrast agents (e.g., fluorescence optical tomography). Here we utalize spatial frequency domain imaging (SFDI), a noninvasive, label-free optical technique, for the wide-field quantification of changes in tissue optical scattering in preclinical tumor models during treatment with chemotherapy and antiangiogenic agents. Optical scattering is particularly sensitive to tissue micro-architectural changes, including those that occur during apoptosis, an early indicator of response to cytotoxicity induced by chemotherapy, thermotherapy, cryotherapy, or radiation therapy. We utilized SFDI to monitor responses of PC3/2G7 prostate tumors and E0771 mammary tumors to treatment with cyclophosphamide or the antiangiogenic agent DC101 for up to 49 days. The SFDI-derived scattering amplitude was highly correlated with cleaved caspase-3, a marker of apoptosis (ρp = 0.75), while the exponent of the scattering wavelength-dependence correlated with the cell proliferation marker PCNA (ρp = 0.69). These optical parameters outperformed tumor volume and several functional parameters (e.g., oxygen saturation and hemoglobin concentration) as an early predictive biomarker of treatment response. Quantitative diffuse optical scattering is thus a promising new early marker of treatment response, which does not require radiation or exogenous contrast agents.

RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues.

In situ analysis of biomarkers is highly desirable in molecular pathology because it allows the examination of biomarker status within the histopathological context of clinical specimens. Immunohistochemistry and DNA in situ hybridization (ISH) are widely used in clinical settings to assess protein and DNA biomarkers, respectively, but clinical use of in situ RNA analysis is rare. This disparity is especially notable when considering the abundance of RNA biomarkers discovered through whole-genome expression profiling. This is largely due to the high degree of technical complexity and insufficient sensitivity and specificity of current RNA ISH techniques. Here, we describe RNAscope, a novel RNA ISH technology with a unique probe design strategy that allows simultaneous signal amplification and background suppression to achieve single-molecule visualization while preserving tissue morphology. RNAscope is compatible with routine formalin-fixed, paraffin-embedded tissue specimens and can use either conventional chromogenic dyes for bright-field microscopy or fluorescent dyes for multiplex analysis. Unlike grind-and-bind RNA analysis methods such as real-time RT-PCR, RNAscope brings the benefits of in situ analysis to RNA biomarkers and may enable rapid development of RNA ISH-based molecular diagnostic assays.

Specific regulation of NRG1 isoform expression by neuronal activity.

Neuregulin 1 (NRG1) is a trophic factor that has been implicated in neural development, neurotransmission, and synaptic plasticity. NRG1 has multiple isoforms that are generated by usage of different promoters and alternative splicing of a single gene. However, little is known about NRG1 isoform composition profile, whether it changes during development, or the underlying mechanisms. We found that each of the six types of NRG1 has a distinct expression pattern in the brain at different ages, resulting in a change in NRG1 isoform composition. In both human and rat, the most dominant are types III and II, followed by either type I or type V, while types IV and VI are the least abundant. The expression of NRG1 isoforms is higher in rat brains at ages of E13 and P5 (in particular type V), suggesting roles in early neural development and in the neonatal critical period. At the cellular level, the majority of NRG1 isoforms (types I, II, and III) are expressed in excitatory neurons, although they are also present in GABAergic neurons and astrocytes. Finally, the expression of each NRG1 isoform is distinctly regulated by neuronal activity, which causes significant increase in type I and IV NRG1 levels. Neuronal activity regulation of type IV expression requires a CRE cis-element in the 5' untranslated region (UTR) that binds to CREB. These results indicate that expression of NRG1 isoforms is regulated by distinct mechanisms, which may contribute to versatile functions of NRG1 and pathologic mechanisms of brain disorders such as schizophrenia.

Drosophila E-cadherin and its binding partner Armadillo/ beta-catenin are required for axonal pathway choices in the developing larval brain.

The fly brain is formed by approximately hundred paired lineages of neurons, each lineage derived from one neuroblast. Embryonic neuroblasts undergo a small number of divisions and produce the primary neurons that form the functioning larval brain. In the larva, neuroblasts produce the secondary lineages that make up the bulk of the adult brain. Axons of a given secondary lineage fasciculate with each other and form a discrete bundle, the secondary axon tract (SAT). Secondary axon tracts prefigure the long axon connections of the adult brain, and therefore pathway choices of SATs made in the larva determine adult brain circuitry. Drosophila Shotgun/E-cadherin (DE-cad) and its binding partner Armadillo/beta-catenin (beta-cat) are expressed in newly born secondary neurons and their axons. The fact that the highly diverse, yet invariant pattern of secondary lineages and SATs has been recently mapped in the wild-type brain enabled us to investigate the role of DE-cad and beta-cat with the help of MARCM clones. Clones were validated by their absence of DE-cad immuno-reactivity. The most significant phenotype consists in the defasciculation and an increased amount of branching of SATs at the neuropile-cortex boundary, as well as subtle changes in the trajectory of SATs within the neuropile. In general, only a fraction of mutant clones in a given lineage showed structural abnormalities. Furthermore, although they all globally express DE-cad and beta-cat, lineages differ in their requirement for DE-cad function. Some lineages never showed morphological abnormalities in MARCM clones, whereas others reacted with abnormal branching and changes in SAT trajectory at a high frequency. We conclude that DE-cad/beta-cat form part of the mechanism that control branching and trajectory of axon tracts in the larval brain.

The Drosophila homologue of the Angelman syndrome ubiquitin ligase regulates the formation of terminal dendritic branches.

Angelman syndrome is a severe neurodevelopmental disorder mostly caused by loss-of-function mutations in the maternal allele of UBE3A, a gene that encodes an E3 ubiquitin ligase. Drosophila UBE3A (dUBE3A) is highly homologous to human UBE3A (hUBE3A) at the amino acid sequence level, suggesting their functional conservation. We generated dUBE3A-null mutant fly lines and found that dUBE3A is not essential for viability. However, loss of dUBE3A activity reduced dendritic branching of sensory neurons in the peripheral nervous system and slowed the growth of terminal dendritic fine processes. Several lines of evidence indicated that dUBE3A regulates dendritic morphogenesis in a cell autonomous manner. Moreover, overexpression of dUBE3A also decreased dendritic branching, suggesting that the proper level of dUBE3A is critically important for the normal dendritic patterning. These findings suggest that dendritic pathology may contribute to neurological deficits in patients with Angelman syndrome.

The steady-state level of the nervous-system-specific microRNA-124a is regulated by dFMR1 in Drosophila.

Fragile X syndrome is the most common form of inherited mental retardation caused by loss of the fragile X mental retardation protein 1 (FMRP). The detailed molecular pathways underlying the pathogenesis of this disorder remain incompletely understood. Here, we show that miR-124a, a nervous-system-specific miRNA, is associated with the Drosophila homolog of FMRP (dFMR1) in vivo. Ectopic expression of wild-type but not mutant miR-124a precursors decreased dendritic branching of dendritic arborization sensory neurons, which was partially rescued by the loss of dFMR1 activity, suggesting that the biogenesis and/or function of miR-124a are partially dependent on dFMR1. Indeed, in contrast with the complete loss of mature miR-124a in Dicer-1 mutants, steady-state levels of endogenous or ectopically expressed mature miR-124a were partially reduced in dfmr1 mutants, whereas the level of pre-miR-124a increased. This effect could be explained in part by the reduced abundance of the Dicer-1-Ago1 complex in the absence of dFMR1. These findings suggest a modulatory role for dFMR1 to maintain proper levels of miRNAs during neuronal development.

Expression profile of the cadherin family in the developing Drosophila brain.

The Drosophila genome encodes 17 members of the cadherin family of adhesion molecules, which in vertebrates has been implicated in patterning the nervous system through cell and axon sorting. With only a few exceptions all cadherins show widespread expression in the larval brain. What expression patterns have in common is that 1) they are global, in the sense that all lineages of the central brain or optic lobe, or both, show expression; and 2) expression is stage-specific: some cadherins are expressed only in primary neurons (located closest to the neuropile), others in early secondary neurons (near the brain surface), or primaries plus late secondaries. The Fat-like cadherins, Fat and Dachsous, as well as Cad96Ca and Cad74A, are expressed in the epithelial optic lobe anlagen, which matches the widespread epithelial expression of these molecules in the embryo. DE-cadherin is restricted to immature secondary neurons and glia; by contrast, DN-cadherin, Flamingo, Cad87A, Cad99C, and Calsyntenin-1 appear in differentiating primary neurons and, at a later stage, some or all secondary neurons. Cad87A is strongly enriched apically in epithelia and in neuronal dendrites. Fat-like, Cad86C, Cad88C, Cad89D, and Dret are expressed ubiquitously in embryonic and larval brains at low or moderate levels. We conclude from this analysis that cadherins are likely to play a role in 'generic' neural functions, such as neurite fasciculation, branching, and synapse formation.

MicroRNA-9a ensures the precise specification of sensory organ precursors in Drosophila.

MicroRNAs (miRNAs) have been implicated in regulating various aspects of animal development, but their functions in neurogenesis are largely unknown. Here we report that loss of miR-9a function in the Drosophila peripheral nervous system leads to ectopic production of sensory organ precursors (SOPs), whereas overexpression of miR-9a results in a severe loss of SOPs. We further demonstrate a strong genetic interaction between miR-9a and senseless (sens) in controlling the formation of SOPs in the adult wing imaginal disc. Moreover, miR-9a suppresses Sens expression through its 3' untranslated region. miR-9a is expressed in epithelial cells, including those adjacent to SOPs within proneural clusters, suggesting that miR-9a normally inhibits neuronal fate in non-SOP cells by down-regulating Sens expression. These results indicate that miR-9a ensures the generation of the precise number of neuronal precursor cells during development.

Regulation of cell adhesion in the Drosophila embryo by phosphorylation of the cadherin-catenin-complex.

Cell-culture studies indicate that tyrosine phosphorylation of the cadherin-catenin-complex (CCC) is one of the post-translational mechanism regulating E-cadherin-mediated cell adhesion. In this investigation, controlled application of a tyrosine phosphatase inhibitor (orthovanadate) and tyrosine kinase inhibitor (tyrphostin) to early Drosophila embryos, followed by biochemical assays and phenotypic analysis, has been utilized to address the mechanism by which tyrosine phosphorylation regulates E-cadherin-mediated cell adhesion in vivo. Our data suggest that, in the Drosophila embryo, beta-catenin (Drosophila homolog Armadillo) is the primary tyrosine-phosphorylated protein in the CCC. The increase in tyrosine phosphorylation correlates with a loss of epithelial integrity and adherens junctions in the ectoderm of early embryos. Late application of the phosphatase inhibitor does not have this effect, presumably because of the formation of septate junctions in late embryos. Co-immunoprecipitation assays have demonstrated that tyrosine hyper-phosphorylation does not cause the dissociation of Drosophila (D)E-cadherin and alpha-catenin or Armadillo, suggesting that abrogation in adhesion is most likely attributable to the detachment of actin-associated proteins from the CCC. Finally, although the Drosophila epidermal growth factor receptor (EGFR), a receptor tyrosine kinase, is linked to the CCC and shows genetic interactions with DE-cadherin, we find that a constitutively active Drosophila EGFR construct does not cause any detectable changes in the level of tyrosine phosphorylation of Armadillo or destabilization of the CCC.

BTB/POZ-zinc finger protein abrupt suppresses dendritic branching in a neuronal subtype-specific and dosage-dependent manner.

How dendrites of different neuronal subtypes exhibit distinct branching patterns during development remains largely unknown. Here we report the mapping and identification of loss-of-function mutations in the abrupt (ab) gene that increased the number of dendritic branches of multiple dendritic (MD) sensory neurons in Drosophila embryos. Ab encodes an evolutionarily conserved transcription factor that contains a BTB/POZ domain and C2H2 zinc finger motifs. We show that ab has a cell-autonomous function in postmitotic neurons to limit dendritic branching. Ab and the homeodomain protein Cut are expressed in distinct but complementary subsets of MD neurons, and Ab functions in a transcriptional program that does not require Cut. Deleting one copy of ab or overexpressing ab had opposite effects on the formation of higher-order dendritic branches, suggesting that the Ab level in a specific neuron directly regulates dendritic complexity. These results demonstrate that dendritic branching can be suppressed by neuronal subtype-specific transcription factors in a cell-autonomous and dosage-dependent manner.

The role of DE-cadherin during cellularization, germ layer formation and early neurogenesis in the Drosophila embryo.

The Drosophila E-cadherin homolog, DE-cadherin, is expressed and required in all epithelial tissues throughout embryogenesis. Due to a strong maternal component of DE-cadherin, its early function during embryogenesis has remained elusive. The expression of a dominant negative DE-cadherin construct (UAS-DE-cad(ex)) using maternally active driver lines allowed us to analyze the requirements for DE-cadherin during this early phase of development. Maternally expressed DE-cad(ex) result in phenotype with variable expressivity. Most severely affected embryos have abnormalities in epithelialization of the blastoderm, resulting in loss of the blastodermal cells' apico-basal polarity and monolayered structure. Another phenotypic class forms a rather normal blastoderm, but shows abnormalities in proliferation and morphogenetic movements during gastrulation and neurulation. Mitosis of the mesoderm occurs prematurely before invagination, and proliferation in the ectoderm, normally a highly ordered process, occurs in a random pattern. Mitotic spindles of ectodermal cells, normally aligned horizontally, frequently occurred vertically or at an oblique angle. This finding further supports recent findings indicating that, in the wild-type ectoderm, the zonula adherens is required for the horizontal orientation of mitotic spindles. Proliferation defects in DE-cad(ex)-expressing embryos are accompanied by the loss of epithelial structure of ectoderm and neuroectoderm. These germ layers form irregular double or triple layers of rounded cells that lack zonula adherens. In the multilayered neuroectoderm, epidermal precursors, neuroblasts and ganglion mother cells occurred intermingled, attesting to the pivotal role of DE-cadherin in delamination and polarized division of neuroblasts. By contrast, the overall number and spacing of neuroblasts was grossly normal, indicating that DE-cadherin-mediated adhesion is less important for cell-cell interaction controlling the ratio of epidermal vs. neural progenitors.

APC-independent activation of NK cells by the Toll-like receptor 3 agonist double-stranded RNA.

Toll-like receptors (TLRs) play a fundamental role in the recognition of bacteria and viruses. TLR3 is activated by viral dsRNA and polyinosinic-polycytidylic acid (poly(I:C)), a synthetic mimetic of viral RNA. We show that NK cells, known for their capacity to eliminate virally infected cells, express TLR3 and up-regulate TLR3 mRNA upon poly(I:C) stimulation. Treatment of highly purified NK cells with poly(I:C) significantly augments NK cell-mediated cytotoxicity. Poly(I:C) stimulation also leads to up-regulation of activation marker CD69 on NK cells. Furthermore, NK cells respond to poly(I:C) by producing proinflammatory cytokines like IL-6 and IL-8, as well as the antiviral cytokine IFN-gamma. The induction of cytokine production by NK cells was preceded by activation of NF-kappaB. We conclude that the ability of NK cells to directly recognize and respond to viral products is important in mounting effective antiviral responses.

Stanniocalcin 1 is an autocrine modulator of endothelial angiogenic responses to hepatocyte growth factor.

Stanniocalcin 1 (STC1) is a secreted glycoprotein originally described as a hormone involved in calcium and phosphate homeostasis in bony fishes. We recently identified the mammalian homolog of this molecule to be highly up-regulated in an in vitro model of angiogenesis, as well as focally and intensely expressed at sites of pathological angiogenesis (e.g. tumor vasculature). In the present study, we report that STC1 is a selective modulator of hepatocyte growth factor (HGF)-induced endothelial migration and morphogenesis, but not proliferation. STC1 did not inhibit proliferative or migratory responses to vascular endothelial growth factor or basic fibroblast growth factor. The mechanism of STC1 inhibitory effects on HGF-induced endothelial migration seem to occur secondary to receptor activation because STC1 did not inhibit HGF-induced c-met receptor phosphorylation, but did block HGF-induced focal adhesion kinase activation. In the mouse femoral artery ligation model of angiogenesis, STC1 expression closely paralleled that of the endothelial marker CD31, and the peak level of STC1 expression occurred after an increase in HGF expression. We propose that STC1 may play a selective modulatory role in angiogenesis, possibly serving as a "stop signal" or stabilizing factor contributing to the maturation of newly formed blood vessels. HGF is a mesenchyme-derived pleiotropic factor with mitogenic, motogenic, and morphogenic activities on a number of different cell types. HGF effects are mediated through a specific tyrosine kinase, c-met, and aberrant HGF and c-met expression are frequently observed in a variety of tumors. Recent studies have shown HGF to be a potent growth factor implicated in wound healing, tissue regeneration, and angiogenesis.

Role of DE-cadherin in neuroblast proliferation, neural morphogenesis, and axon tract formation in Drosophila larval brain development.

In the wild-type brain, the Drosophila classic cadherin DE-cadherin is expressed globally by postembryonic neuroblasts and their lineages ("secondary lineages"), as well as glial cells. To address the role of DE-cadherin in the larval brain, we took advantage of the dominant-negative DE-cad(ex) construct, the expression of which was directed to neurons, glial cells, or both. Global expression of DE-cad(ex) driven by a heat pulse during the early second instar resulted in a severe phenotype that included deficits in neural proliferation. Neuroblasts appeared in approximately normal numbers but had highly reduced mitotic activity. When the DE-cad(ex) construct was driven by the glial-specific driver gcm-Gal4, the effect of DE-cad(ex) on neuroblast proliferation could be replicated, which indicates that DE-cadherin acts in glial cells to promote proliferation of neuroblasts. Expression of DE-cad(ex) in neurons, cortex glia, or both results in abnormalities in cortex layering and in trajectories of secondary axons. In the wild-type brain, neuroblasts and neurons generated at different time points are arranged concentrically around the neuropile, with the DE-cadherin-positive neuroblasts and young secondary neurons at the surface, followed by older secondary neurons and primary neurons. Axons of secondary lineages follow a straight radial course toward the neuropile. Processes of glial cells located in the cortex form a scaffold, called trophospongium, that enwraps neuroblasts and neurons. Expression of DE-cad(ex) in neurons, cortex glia, or both disrupted the regular placement of neuroblasts and secondary neurons and resulted in abnormal trajectories of cell body fiber tracts. We conclude that DE-cadherin plays a pivotal role in larval brain proliferation, brain cortex morphogenesis, and axon growth.

Early development of the Drosophila brain: V. Pattern of postembryonic neuronal lineages expressing DE-cadherin.

The Drosophila E-cadherin homolog, DE-cadherin, is expressed postembryonically by brain neuroblasts and their lineages of neurons ("secondary lineages"). DE-cadherin appears in neuroblasts as soon as they can be identified by their increase in size and then remains expressed uninterruptedly throughout larval life. DE-cadherin remains transiently expressed in the cell bodies and axons of neurons produced by neuroblast proliferation. In general, axons of neurons belonging to one lineage form tight bundles. The trajectories of these bundles are correlated with the location of the neuronal lineages to which they belong. Thus, axon bundles of lineages that are neighbors in the cortex travel parallel to each other and reach the neuropile at similar positions. It is, therefore, possible to assign coherent groups of neuroblasts and their lineages to the individual neuropile compartments and long axon tracts introduced in the accompanying articles (Nassif et al. [2003] J Comp Neurol 455:417-434; Younossi-Hartenstein et al. [2003] J Comp Neurol 455:435-450). In this study, we have reconstructed the pattern of secondary lineages and their projection in relationship to the compartments and Fasciclin II-positive long axon tracts. Based on topology and axonal trajectory, the lineages of the central brain can be subdivided into 11 groups that can be followed throughout successive larval stages. The map of larval lineages and their axonal projection will be important for future studies on postembryonic neurogenesis in Drosophila. It also lays a groundwork for investigating the role of DE-cadherin in larval brain development.

Time course of skeletal muscle repair and gene expression following acute hind limb ischemia in mice.

DNA microarrays were used to measure the time course of gene expression during skeletal muscle damage and regeneration in mice following femoral artery ligation (FAL). We found 1,289 known sequences were differentially expressed between the FAL and control groups. Gene expression peaked on day 3, and the functional cluster "inflammation" contained the greatest number of genes. Muscle function was depressed for 3 days postligation, but returned to normal by day 7. Decreased muscle function was accompanied by reduced expression of genes involved in mitochondrial energy production, muscle contraction, and calcium handling. The induction of MyoD on day 1 denoted the beginning of muscle regeneration and was followed by the reemergence of the embryonic forms of muscle contractile proteins, which peaked at day 7. Transcriptional analysis indicated that the ischemic skeletal muscle may transition through a functional adaptation stage with recovery of contractile force prior to full regeneration. Several members of the insulin-like growth factor axis were coordinately induced in a time frame consistent with their playing a role in the regenerative process.

Interaction between EGFR signaling and DE-cadherin during nervous system morphogenesis.

Dynamically regulated cell adhesion plays an important role during animal morphogenesis. Here we use the formation of the visual system in Drosophila embryos as a model system to investigate the function of the Drosophila classic cadherin, DE-cadherin, which is encoded by the shotgun (shg) gene. The visual system is derived from the optic placode which normally invaginates from the surface ectoderm of the embryo and gives rise to two separate structures, the larval eye (Bolwig's organ) and the optic lobe. The optic placode dissociates and undergoes apoptotic cell death in the absence of DE-cadherin, whereas overexpression of DE-cadherin results in the failure of optic placode cells to invaginate and of Bolwig's organ precursors to separate from the placode. These findings indicate that dynamically regulated levels of DE-cadherin are essential for normal optic placode development. It was shown previously that overexpression of DE-cadherin can disrupt Wingless signaling through titration of Armadillo out of the cytoplasm to the membrane. However, the observed defects are likely the consequence of altered DE-cadherin mediated adhesion rather than a result of compromising Wingless signaling, as overexpression of a DE-cadherin-alpha-catenin fusion protein, which lacks Armadillo binding sites, causes similar defects as DE-cadherin overexpression. We further studied the genetic interaction between DE-cadherin and the Drosophila EGF receptor homolog, EGFR. If EGFR function is eliminated, optic placode defects resemble those following DE-cadherin overexpression, which suggests that loss of EGFR results in an increased adhesion of optic placode cells. An interaction between EGFR and DE-cadherin is further supported by the finding that expression of a constitutively active EGFR enhances the phenotype of a weak shg mutation, whereas a mutation in rhomboid (rho) (an activator of the EGFR ligand Spitz) partially suppresses the shg mutant phenotype. Finally, EGFR can be co-immunoprecipitated with anti-DE-cadherin and anti-Armadillo antibodies from embryonic protein extracts. We propose that EGFR signaling plays a role in morphogenesis by modulating cell adhesion.