Multi-level information fusion for environment aware robotic navigation

This thesis develops the hardware and software framework for an integrated navigation system. Dynamic data fusion algorithms are used to develop a system with a high level of resistance to the typical problems that affect standard navigation systems.

Transcription and expression of Plasmodium falciparum histidine-rich proteins in different stages and strains : implications for rapid diagnostic tests

Background: Although rapid diagnostic tests (RDTs) for Plasmodium falciparum infection that target histidine rich protein 2 (PfHRP2) are generally sensitive, their performance has been reported to be variable. One possible explanation for variable test performance is differences in expression level of PfHRP in different parasite isolates. Methods: Total RNA and protein were extracted from synchronised cultures of 7 P. falciparum lines over 5 time points of the life cycle, and from synchronised ring stages of 10 falciparum lines. Using quantitative real-time polymerase chain reaction, Western blot analysis and ELISA we investigated variations in the transcription and protein levels of pfhrp2, pfhrp3 and PfHRP respectively in the different parasite lines, over the parasite intraerythrocytic life cycle. Results: Transcription of pfhrp2 and pfhrp3 in different parasite lines over the parasite life cycle was observed to vary relative to the control parasite K1. In some parasite lines very low transcription of these genes was observed. The peak transcription was observed in ring-stage parasites. Pfhrp2 transcription was observed to be consistently higher than pfhrp3 transcription within parasite lines. The intraerythrocytic lifecycle stage at which the peak level of protein was present varied across strains. Total protein levels were more constant relative to total mRNA transcription, however a maximum 24 fold difference in expression at ring-stage parasites relative to the K1 strain was observed. Conclusions: The levels of transcription of pfhrp2 and pfhrp3, and protein expression of PfHRP varied between different P. falciparum strains. This variation may impact on the detection sensitivity of PfHRP2-detecting RDTs.

Sequential decision fusion of multibiometrics applied to text-dependent speaker verification for controlled errors

Reliability of the performance of biometric identity verification systems remains a significant challenge. Individual biometric samples of the same person (identity class) are not identical at each presentation and performance degradation arises from intra-class variability and inter-class similarity. These limitations lead to false accepts and false rejects that are dependent. It is therefore difficult to reduce the rate of one type of error without increasing the other. The focus of this dissertation is to investigate a method based on classifier fusion techniques to better control the trade-off between the verification errors using text-dependent speaker verification as the test platform. A sequential classifier fusion architecture that integrates multi-instance and multisample fusion schemes is proposed. This fusion method enables a controlled trade-off between false alarms and false rejects. For statistically independent classifier decisions, analytical expressions for each type of verification error are derived using base classifier performances. As this assumption may not be always valid, these expressions are modified to incorporate the correlation between statistically dependent decisions from clients and impostors. The architecture is empirically evaluated by applying the proposed architecture for text dependent speaker verification using the Hidden Markov Model based digit dependent speaker models in each stage with multiple attempts for each digit utterance. The trade-off between the verification errors is controlled using the parameters, number of decision stages (instances) and the number of attempts at each decision stage (samples), fine-tuned on evaluation/tune set. The statistical validation of the derived expressions for error estimates is evaluated on test data. The performance of the sequential method is further demonstrated to depend on the order of the combination of digits (instances) and the nature of repetitive attempts (samples). The false rejection and false acceptance rates for proposed fusion are estimated using the base classifier performances, the variance in correlation between classifier decisions and the sequence of classifiers with favourable dependence selected using the 'Sequential Error Ratio' criteria. The error rates are better estimated by incorporating user-dependent (such as speaker-dependent thresholds and speaker-specific digit combinations) and class-dependent (such as clientimpostor dependent favourable combinations and class-error based threshold estimation) information. The proposed architecture is desirable in most of the speaker verification applications such as remote authentication, telephone and internet shopping applications. The tuning of parameters - the number of instances and samples - serve both the security and user convenience requirements of speaker-specific verification. The architecture investigated here is applicable to verification using other biometric modalities such as handwriting, fingerprints and key strokes.

The role of immunoglobulins and their transporters in urogenital Chlamydial infections

Chlamydia trachomatis infections of the male and female reproductive tracts are the world's leading sexually transmitted bacterial disease, and can lead to damaging pathology, scarring and infertility. The resolution of chlamydial infection requires the development of adaptive immune responses to infection, and includes cell-mediated and humoral immunity. Whilst cluster of differentiation (CD)4+ T cells are known to be essential in clearance of infection [1], they are also associated with immune cell infiltration, autoimmunity and infertility in the testes [2-3]. Conversely, antibodies are less associated with inflammation, are readily transported into the reproductive tracts, and can offer lumenal neutralization of chlamydiae prior to infection. Antibodies, or immunoglobulins (Ig), play a supportive role in the resolution of chlamydial infections, and this thesis sought to define the function of IgA and IgG, against a variety of chlamydial antigens expressed during the intracellular and extracellular stages of the chlamydial developmental cycle. Transport of IgA and IgG into the mucosal lumen is facilitated by receptor-mediated transcytosis yet the expression profile (under normal conditions and during urogenital chlamydial infection) of the polymeric immunoglobulin receptor (pIgR) and the neonatal Fc receptor (FcRn) remains unknown. The expression profile of pIgR and FcRn in the murine male reproductive tract was found to be polarized to the lower and upper reproductive tract tissues respectively. This demonstrates that the two receptors have a tissue tropism, which must be considered when targeting pathogens that colonize different sites. In contrast, the expression of pIgR and FcRn in the female mouse was found to be distributed in both the upper and lower reproductive tracts. When urogenitally infected with Chlamydia muridarum, both male and female reproductive tracts up-regulated expression of pIgR and down-regulated expression of FcRn. Unsurprisingly, the up-regulation of pIgR increased the concentration of IgA in the lumen. However, down-regulation of FcRn, prevented IgG uptake and led to an increase or pooling of IgG in lumenal secretions. As previous studies have identified the importance of pIgR-mediated delivery of IgA, as well as the potential of IgA to bind and neutralize intracellular pathogens, IgA against a variety of chlamydial antigens was investigated. The protection afforded by IgA against the extracellular antigen major outer membrane protein (MOMP), was found to be dependent on pIgR expression in vitro and in vivo. It was also found that in the absence of pIgR, no protection was afforded to mice previously immunized with MOMP. The protection afforded from polyclonal IgA against the intracellular chlamydial antigens; inclusion membrane protein A (IncA), inclusion membrane proteins (IncMem) and secreted chlamydial protease-like activity factor (CPAF) were produced and investigated in vitro. Antigen-specific intracellular IgA was found to bind to the respective antigen within the infected cell, but did not significantly reduce inclusion formation (p > 0.05). This suggests that whilst IgA specific for the selected antigens was transported by pIgR to the chlamydial inclusion, it was unable to prevent growth. Similarly, immunization of male mice with intracellular chlamydial antigens (IncA or IncMem), followed by depletion CD4+ T cells, and subsequent urogenital C. muridarum challenge, provided minimal pIgR-mediated protection. Wild type male mice immunized with IncA showed a 57 % reduction (p < 0.05), and mice deficient in pIgR showed a 35 % reduction (p < 0.05) in reproductive tract chlamydial burden compared to control antigen, and in the absence of CD4+ T cells. This suggests that pIgR and secretory IgA (SIgA) were playing a protective role (21 % pIgR-mediated) in unison with another antigen-specific immune mechanism (36 %). Interestingly, IgA generated during a primary respiratory C. muridarum infection did not provide a significant amount of protection to secondary urogenital C. muridarum challenge. Together, these data suggest that IgA specific for an extracellular antigen (MOMP) can play a strong protective role in chlamydial infections, and that IgA targeting intracellular antigens is also effective but dependent on pIgR expression in tissues. However, whilst not investigated here, IgA targeting and blocking other intracellular chlamydial antigens, that are more essential for replication or type III secretion, may be more efficacious in subunit vaccines. Recently, studies have demonstrated that IgG can neutralize influenza virus by trafficking IgG-bound virus to lysosomes [4]. We sought to determine if this process could also traffic chlamydial antigens for degradation by lysosomes, despite Chlamydia spp. actively inhibiting fusion with the host endocytic pathway. As observed in pIgR-mediated delivery of anti-IncA IgA, FcRn similarly transported IgG specific for IncA which bound the inclusion membrane. Interestingly, FcRn-mediated delivery of anti-IncA IgG significantly decreased inclusion formation by 36 % (p < 0.01), and induced aberrant inclusion morphology. This suggests that unlike IgA, IgG can facilitate additional host cellular responses which affect the intracellular niche of chlamydial growth. Fluorescence microscopy revealed that IgG also bound the inclusion, but unlike influenza studies, did not induce the recruitment of lysosomes. Notably, anti-IncA IgG recruited sequestosomes to the inclusion membrane, markers of the ubiquitin/proteasome pathway and major histocompatibility complex (MHC) class I loading. To determine if the protection against C. muridarum infection afforded by IncA IgG in vitro translated in vivo, wild type mice and mice deficient in functional FcRn and MHC-I, were immunized, depleted of CD4+, and urogenitally infected with C. muridarum. Unlike in pIgR-deficient mice, the protection afforded from IncA immunization was completely abrogated in mice lacking functional FcRn and MHC-I/CD8+. Thus, both anti-IncA IgA and IgG can bind the inclusion in a pIgR and FcRn-mediated manner, respectively. However, only IgG mediates a higher reduction in chlamydial infection in vitro and in vivo suggesting more than steric blocking of IncA had occurred. Unlike anti-MOMP IgA, which reduced chlamydial infection of epithelial cells and male mouse tissues, IgG was found to enhance infectivity in vitro, and in vivo. Opsonization of EBs with MOMP-IgG enhanced inclusion formation of epithelial cells in a MOMP-IgG dose-dependent and FcRn-dependent manner. When MOMP-IgG opsonized EBs were inoculated into the vagina of female mice, a small but non-significant (p > 0.05) enhancement of cervicovaginal C. muridarum shedding was observed three days post infection in mice with functional FcRn. Interestingly, infection with opsonized EBs reduced the intensity of the peak of infection (day six) but protracted the duration of infection by 60 % in wild type mice only. Infection with EBs opsonized in IgG also significantly increased (p < 0.05) hydrosalpinx formation in the oviducts and induced lymphocyte infiltration uterine horns. As MOMP is an immunodominant antigen, and is widely used in vaccines, the ability of IgG specific to extracellular chlamydial antigens to enhance infection and induce pathology needs to be considered. Together, these data suggest that immunoglobulins play a dichotomous role in chlamydial infections, and are dependent on antigen specificity, FcRn and pIgR expression. FcRn was found to be highly expressed in upper male reproductive tract, whilst pIgR was dominantly expressed in the lower reproductive tract. Conversely, female mice expressed FcRn and pIgR in both the lower and upper reproductive tracts. In response to a normal chlamydial infection, pIgR is up-regulated increasing secretory IgA release, but FcRn is down-regulated preventing IgG uptake. Similarly to other studies [5-6], we demonstrate that IgA and IgG generated during primary chlamydial infections plays a minor role in recall immunity, and that antigen-specific subunit vaccines can offer more protection. We also show that both IgA and IgG can be used to target intracellular chlamydial antigens, but that IgG is more effective. Finally, IgA against the extracellular antigen MOMP can afford protection, whist IgG plays a deleterious role by increasing infectivity and inducing damaging immunopathology. Further investigations with additional antigens or combination subunit vaccines will enhance our understanding the protection afforded by antibodies against intracellular and extracellular pathogenic antigens, and help improve the development of an efficacious chlamydial vaccine.

Analysis of acute-phase proteins, AHSG, C3, CLI, HP and SAA, reveals distinctive expression patterns associated with breast, colorectal and lung cancer

Early detection, clinical management and disease recurrence monitoring are critical areas in cancer treatment in which specific biomarker panels are likely to be very important in each of these key areas. We have previously demonstrated that levels of alpha-2-heremans-schmid-glycoprotein (AHSG), complement component C3 (C3), clusterin (CLI), haptoglobin (HP) and serum amyloid A (SAA) are significantly altered in serum from patients with squamous cell carcinoma of the lung. Here, we report the abundance levels for these proteins in serum samples from patients with advanced breast cancer, colorectal cancer (CRC) and lung cancer compared to healthy controls (age and gender matched) using commercially available enzyme-linked immunosorbent assay kits. Logistic regression (LR) models were fitted to the resulting data, and the classification ability of the proteins was evaluated using receiver-operating characteristic curve and leave-one-out cross-validation (LOOCV). The most accurate individual candidate biomarkers were C3 for breast cancer [area under the curve (AUC) = 0.89, LOOCV = 73%], CLI for CRC (AUC = 0.98, LOOCV = 90%), HP for small cell lung carcinoma (AUC = 0.97, LOOCV = 88%), C3 for lung adenocarcinoma (AUC = 0.94, LOOCV = 89%) and HP for squamous cell carcinoma of the lung (AUC = 0.94, LOOCV = 87%). The best dual combination of biomarkers using LR analysis were found to be AHSG + C3 (AUC = 0.91, LOOCV = 83%) for breast cancer, CLI + HP (AUC = 0.98, LOOCV = 92%) for CRC, C3 + SAA (AUC = 0.97, LOOCV = 91%) for small cell lung carcinoma and HP + SAA for both adenocarcinoma (AUC = 0.98, LOOCV = 96%) and squamous cell carcinoma of the lung (AUC = 0.98, LOOCV = 84%). The high AUC values reported here indicated that these candidate biomarkers have the potential to discriminate accurately between control and cancer groups both individually and in combination with other proteins. Copyright © 2011 UICC.

bcl-2 and c-erbB-2 proteins are involved in the regulation of VEGF and of thymidine phosphorylase angiogenic activity in non-small-cell lung cancer

Tumour angiogenesis has been recently recognised as one of the most important prognostic factors in lung cancer. Although a variety of angiogenic factors have been identified, the angiogenesis process remains poorly understood. Bcl-2, c-erbB-2 and p53 are well-known oncogenes involved in non- small-cell lung cancer pathogenesis. A direct correlation of thymidine phosphorylase (TP) and of vascular endothelial growth factor (VEGF) with intratumoural angiogenesis has been reported. In the present study we investigated the possible regulatory role if bcl-2, c-erB-2 proteins in angiogenesis and in VEGF and TP expression in non-small-cell lung cancer. Two hundred sixteen specimens from T1,2-NO, 1 staged patients treated with surgery alone were immunohistochemically examined. Bcl-2 and c-erbB-2 were significantly inversely related to each other (P = 0.04) and both were inversely associated with microvessel density (P < 0.02). High TP and VEGF reactivity was statistically related to loss of bcl-2 expression (P < 0.01). A significant co-expression of c-erbB-2 with TP was noted (P = 0.01). However, TP expression was related to high angiogenesis only in cases with absence of c-erB-2 expression (P < 0.0001). c-erbB-2 expression in poorly vascularised tumours was linked with poor outcome (P = 0.03). The present study provides strong evidence that the bcl-2 gene has a suppressive function over genes involved in both angiogenesis (VEGF and TP) and cell migration (c- erbB-2) in NSCLC. TP and c-erbB-2 proteins are significantly, and often simultaneously, expressed in bcl-2 negative cases. However, expression of the c-erbB-2 abolishes the TP-related angiogenic activity. Whether this is a result of a direct activity of the c-erbB-2 protein or a consequence of a c- erbB-2-related immune response remains to be further investigated.

Quantitative proteomic approach to study subcellular localization of membrane proteins

As proteins within cells are spatially organized according to their role, knowledge about protein localization gives insight into protein function. Here, we describe the LOPIT technique (localization of organelle proteins by isotope tagging) developed for the simultaneous and confident determination of the steady-state distribution of hundreds of integral membrane proteins within organelles. The technique uses a partial membrane fractionation strategy in conjunction with quantitative proteomics. Localization of proteins is achieved by measuring their distribution pattern across the density gradient using amine-reactive isotope tagging and comparing these patterns with those of known organelle residents. LOPIT relies on the assumption that proteins belonging to the same organelle will co-fractionate. Multivariate statistical tools are then used to group proteins according to the similarities in their distributions, and hence localization without complete centrifugal separation is achieved. The protocol requires approximately 3 weeks to complete and can be applied in a high-throughput manner to material from many varied sources.

Sub-cellular localization of membrane proteins

In eukaryotes, numerous complex sub-cellular structures exist. The majority of these are delineated by membranes. Many proteins are trafficked to these in order to be able to carry out their correct physiological function. Assigning the sub-cellular location of a protein is of paramount importance to biologists in the elucidation of its role and in the refinement of knowledge of cellular processes by tracing certain activities to specific organelles. Membrane proteins are a key set of proteins as these form part of the boundary of the organelles and represent many important functions such as transporters, receptors, and trafficking. They are, however, some of the most challenging proteins to work with due to poor solubility, a wide concentration range within the cell and inaccessibility to many of the tools employed in proteomics studies. This review focuses on membrane proteins with particular emphasis on sub-cellular localization in terms of methodologies that can be used to determine the accurate location of membrane proteins to organelles. We also discuss what is known about the membrane protein cohorts of major organelles.

The roles of plant dsRNA-binding proteins in RNAi-like pathways

Dicers are associated with double-stranded RNA-binding proteins (dsRBPs) in animals. In the plant, Arabidopsis, there are four dicer-like (DCL) proteins and five potential dsRBPs. These DCLs act redundantly and hierarchically. However, we show there is little or no redundancy or hierarchy amongst the DRBs in their DCL interactions. DCL1 operates exclusively with DRB1 to produce micro (mi)RNAs, DCL4 operates exclusively with DRB4 to produce trans-acting (ta) siRNAs and 21nt siRNAs from viral RNA. DCL2 and DCL3 produce viral siRNAs without requiring assistance from any dsRBP. DRB2, DRB3 and DRB5 appear unnecessary for mi-, tasi-, viral si-, or heterochromatinising siRNA production but act redundantly in a developmental pathway. © 2008 Federation of European Biochemical Societies.

Biopharming the SimpliRED™ HIV diagnostic reagent in barley, potato and tobacco

This is the first report of an antibody-fusion protein expressed in transgenic plants for direct use in a medical diagnostic assay. By the use of gene constructs with appropriate promoters, high level expression of an anti-glycophorin single-chain antibody fused to an epitope of the HIV virus was obtained in the leaves and stems of tobacco, tubers of potato and seed of barley. This fusion protein replaces the SimpliRED™ diagnostic reagent, used for detecting the presence of HIV-1 antibodies in human blood. The reagent is expensive and laborious to produce by conventional means since chemical modifications to a monoclonal antibody are required. The plant-produced fusion protein was fully functional (by ELISA) in crude extracts and, for tobacco at least, could be used without further purification in the HIV agglutination assay. All three crop species produced sufficient reagent levels to be superior bioreactors to bacteria or mice, however barley grain was the most attractive bioreactor as it expressed the highest level (150 μg of reagent g-1), is inexpensive to produce and harvest, poses a minuscule gene flow problem in the field, and the activity of the reagent is largely undiminished in stored grain. This work suggests that barley seed will be an ideal factory for the production of antibodies, diagnostic immunoreagents, vaccines and other pharmaceutical proteins.

Genome segment 5 of rice ragged stunt virus encodes a virion protein

The complete nucleotide sequence of the genome segment 5 (S5) of a Thai isolate of rice ragged stunt virus (RRSV) was determined. The 2682 nucleotide sequence contains a single long open reading frame capable of encoding a polypeptide with a molecular mass of ~91 kDa. Polypeptides encoded by various truncated cDNAs of S5 were expressed using the pGEX fusion protein vector and the highest level of fusion protein was obtained from a construct encoding a hydrophilic region of S5 protein. Antibodies raised against this fusion protein recognized a minor polypeptide, with a molecular mass of ~ 91 kDa, that was present in purified preparations of RRSV particles, infected insect vectors and infected rice plants. This indicates that RRSV S5 encodes a minor structural protein. Comparing the RRSV S5 sequence with sequences of other reo-viruses did not reveal any significant sequence similarities.

Brain-inspired sensor fusion for navigating robots

Current state of the art robot mapping and navigation systems produce impressive performance under a narrow range of robot platform, sensor and environmental conditions, in contrast to animals such as rats that produce “good enough” maps that enable them to function under an incredible range of situations. In this paper we present a rat-inspired featureless sensor-fusion system that assesses the usefulness of multiple sensor modalities based on their utility and coherence for place recognition, without knowledge as to the type of sensor. We demonstrate the system on a Pioneer robot in indoor and outdoor environments with abrupt lighting changes. Through dynamic weighting of the sensors, the system is able to perform correct place recognition and mapping where the static sensor weighting approach fails.

Decision fusion from parts and samples for robust iris recognition

Fusion techniques can be used in biometrics to achieve higher accuracy. When biometric systems are in operation and the threat level changes, controlling the trade-off between detection error rates can reduce the impact of an attack. In a fused system, varying a single threshold does not allow this to be achieved, but systematic adjustment of a set of parameters does. In this paper, fused decisions from a multi-part, multi-sample sequential architecture are investigated for that purpose in an iris recognition system. A specific implementation of the multi-part architecture is proposed and the effect of the number of parts and samples in the resultant detection error rate is analysed. The effectiveness of the proposed architecture is then evaluated under two specific cases of obfuscation attack: miosis and mydriasis. Results show that robustness to such obfuscation attacks is achieved, since lower error rates than in the case of the non-fused base system are obtained.

Laser-radar data fusion with Gaussian Process Implicit Surfaces

This work considers the problem of building high-fidelity 3D representations of the environment from sensor data acquired by mobile robots. Multi-sensor data fusion allows for more complete and accurate representations, and for more reliable perception, especially when different sensing modalities are used. In this paper, we propose a thorough experimental analysis of the performance of 3D surface reconstruction from laser and mm-wave radar data using Gaussian Process Implicit Surfaces (GPIS), in a realistic field robotics scenario. We first analyse the performance of GPIS using raw laser data alone and raw radar data alone, respectively, with different choices of covariance matrices and different resolutions of the input data. We then evaluate and compare the performance of two different GPIS fusion approaches. The first, state-of-the-art approach directly fuses raw data from laser and radar. The alternative approach proposed in this paper first computes an initial estimate of the surface from each single source of data, and then fuses these two estimates. We show that this method outperforms the state of the art, especially in situations where the sensors react differently to the targets they perceive.

The role of the Eph and ephrin proteins in prostate cancer

Prostate cancer is the most commonly diagnosed malignancy and the second leading cause of cancer related deaths in Australian men. Treatment in the early stages of the disease involves surgery, radiation and/or hormone therapy. However, in late stages of the disease these treatments are no longer effective and only palliative care is available. Therefore, there is a focus on exploration of novel therapies to increase survival and treatment efficacy. Advanced prostate cancer is characterised by bone or other distant metastasis. Spreading of the primary tumour to a secondary location is a complex process requiring an initial loss in cell-cell adhesion followed by increased cell migration and invasion. One gene family that has been known to affect cell-to-cell contact in other model systems are the Eph receptor tyrosine kinases. They are the largest family of receptor tyrosine kinases made up of 14 vertebrate Eph receptors that bind to nine cell membrane bound ephrin ligands. Eph-ephrin interaction is crucial in regulating cell behaviour in developmental processes and it is now thought that the underlying mechanisms involved in development may also be involved in cancer. Aberrant expression has been reported in many human malignancies including prostate cancer. Furthermore, expression has been linked with metastasis and poor prognosis in other tumour models. This study explores the potential role of the Eph receptor family in prostate cancer, in particular the roles of EphA2, EphA3 and ephrin-A5. Gene expression profiles were established for the Eph family in a series of prostate cancer cell lines using quantitative real time RT-PCR. A smaller subset of the most prominently expressed genes was chosen to screen a cohort of clinical samples. Elevated levels of EphA2, EphA3 and their ligands, ephrin-A1 and ephrin-A5 were observed in individual cell lines. Interestingly high EphA3 expression was observed in the androgen responsive cell lines while EphA2 was more prominent in the androgen independent cell lines. However, studies using 5-dihydrotestosterone suggest that EphA3 expression in not regulated by androgen. Cells expressing EphA2 showed a greater ability for migration and invasion while cells expressing EphA3 showed poor migration and invasion. Forced expression of EphA2 in the LNCaP cell line resulted in a more invasive phenotype while forced expression of EphA3 in the PC-3 cell line suggests a possible negative effect for EphA3 on cell migration and invasion. Cell signalling studies show activation of EphA2 decreases activity of proteins thought to be involved in pathways regulating cell movement including Akt, Src and FAK. Changes to the activation status of Rho family members, including RhoA and Rac1, associated with reorganisation of the actin cytoskeleton, an important part of cell migration was also observed. As a result, activation of EphA2 in PC-3 cells resulted in a less invasive phenotype. A novel finding in this study was the discovery of a combination of two EphA2 Mabs able to activate EphA2. Preliminary results show a potential for this antibody combination to reduce prostate cancer invasion in vitro. A unique aspect of Eph-ephrin interaction is the resulting bi-directional signalling that occurs through both the receptor and ligand. In this study a potential role for ephrin-A5 mediated signalling in prostate cancer was observed. LNCaP cells express high levels of EphA3 and its high affinity ligand ephrin-A5. In stripe assays, used to study guidance cues, LNCaP cells show strong attraction/migration to EphA3-Fc stripes but not ephrin-A5-Fc stripes suggesting ephrin-A5 mediated reverse cell signalling is involved. Knockdown of ephrin-A5 using shRNA resulted in a decrease in attraction/migration to EphA3-Fc stripes. Furthermore a reduction in proliferation was also observed in vitro. A subcutaneous xenograft model using ephrin-A5 shRNA cells versus controls showed a decrease in tumour formation. This study demonstrates a difference in EphA2 and EphA3 function in prostate cancer migration/invasion and a potential role for ephrin-A5 in prostate cancer cell adhesion and growth.