Search

Abstract Reading typically involves phonological mediation, especially for transparent orthographies with a regular letter to sound correspondence. In this study we ask whether phonological coding is a necessary part of the reading process by examining prelingually deaf individuals who are skilled readers of Spanish. We conducted two EEG experiments exploiting the pseudohomophone effect, in which nonwords that sound like words elicit phonological encoding during reading. The first, a semantic categorization task with masked priming, resulted in modulation of the N250 by pseudohomophone primes in hearing but not in deaf readers. The second, a lexical decision task, confirmed the pattern: hearing readers had increased errors and an attenuated N400 response for pseudohomophones compared to control pseudowords, whereas deaf readers did not treat pseudohomophones any differently from pseudowords, either behaviourally or in the ERP response. These results offer converging evidence that skilled deaf readers do not rely on phonological coding during visual word recognition. Furthermore, the finding demonstrates that reading can take place in the absence of phonological activation, and we speculate about the alternative mechanisms that allow these deaf individuals to read competently.

There is strong published and unpublished evidence that our CD105 Mab E9, which is highly reactive with angiogenic endothelial cells, could be a useful reagent to target the vasculature of solid tumors in man. Since Mab E9 does not cross-react with animal tissues, we undertook here to evaluate its localization using human kidney as an ex vivo model. Perfusion was performed through the renal artery of 99Tcm-labeled purified CD105 Mab in freshly excised kidneys from 7 patients with renal carcinoma. In all 7 cases, immunoscintigraphs showed the presence of well-defined radioactive hot spots, which matched the positions of the tumors as identified by presurgery MRI scans and subsequent histopathologic examination. Importantly, in one instance, where a presurgery MRI scan had identified only one tumor, immunoscintigraphs showed 2 distinct hot spots of radioactivity. The pathology report confirmed that the additional hot spot corresponded to a small secondary well-vascularized tumor. The implication of this finding is that the radiolabeled Mab, E9, may be of use in the detection of metastatic disease. That the labeling of tumors was specific was confirmed when prior perfusion of unlabeled mab E9 in 2 kidneys completely blocked the localization of 99Tcm-conjugated Mab E9. Radioactivity in samples of tumor and normal tissue taken from 7 kidneys was counted in a gamma counter. In all cases, there was a greater uptake of radioactivity in tumors compared with the corresponding normal kidneys. The median values, adjusted per gram wet weight, for 99Tcm were 14.8 times (range, 4.8–113.0) greater in kidney tumors than in normal kidney tissue (p < 0.007). Immunofluorescent staining of cryostat sections of tumor tissues in each of the 7 cases showed strong and uniform localization of Mab E9 in tumor microvessels. Interestingly, chimeric staining of endothelial cells (ECs) was seen in an occasional microvessel segment. That is, while most of the ECs lining a microvessel were strongly stained, an occasional EC was negative. This was not an artifact of staining. Unstained ECs may be nonangiogenic or apoptotic since CD105 is a proliferation/activation-associated antigen. Further investigations are warranted to establish the pharmacokinetics of 99Tcm-labeled CD105 antibody in vivo. This would enable us to determine whether an apparently highly successful ex vivo study has the potential for tumor imaging/therapeutic vascular targeting in patients with cancer. © 2004 Wiley-Liss, Inc. [ABSTRACT FROM AUTHOR]