Identification of a macromolecule containing an anticarcinoembryonic antigen-reactive substance and immunoglobulin M in human pancreatic cancer.

Ascitic fluid from a patient with carcinoma of the pancreas was fractionated by ammonium sulfate precipitation. The fraction precipitated between 25 and 50% saturation of ammonium sulfate was sequentially chromatographed on Sephadex G-200 and Sepharose 6B. A macromolecular fraction (greater than 10(6) daltons) obtained was found to react with both antihuman IgM and antiserum to carcinoembryonic antigen (CEA). This fraction was further purified by adsorption with protein A-Sepharose CL-4B and chromatography on DEAE-Sephacel. The purified macromolecular fraction had a sedimentation value of 28S as determined by ultracentrifugation. Upon dissociation of the purified macromolecule at pH 2.3 and purification of the dissociated components on Sepharose CL-2B and BioGel A 1.5M, a 19S protein and a 5S protein were recovered. The 19S protein showed a complete line of identity with a reference human IgM when reacted with antihuman IgM in gel diffusion, whereas the 5S protein showed a partial immunologic identity with colon CEA against anti-CEA. These results indicated the existence of an IgM-containing macromolecular complex with an anti-CEA cross-reactive substance in the extracellular fluid of human pancreatic cancer.

Purification and characterization of a human pancreas-specific antigen.

A pancreas-specific antigen was identified by immunologic techniques and purified from saline extract of human pancreas. The purified pancreas-specific antigen was shown to be homogeneous by polyacrylamide gel electrophoresis under both denaturing and non-denaturing conditions. It had a molecular weight of 44000 as estimated by gel filtration or sodium dodecyl sulfate-gel electrophoresis, and a sedimentation coefficient of 3.4 S as analyzed by sucrose gradient centrifugation. Pancreas-specific antigen possessed an isoelectric point of 4.9 and migrated to alpha-beta region upon immunoelectrophoresis. By colorimetric assay procedures, pancreas-specific antigen exhibited no enzyme activity, such as amylase, protease, esterase, lipase, acid phosphatase, alkaline phosphatase peroxidase, deoxyribonuclease or ribonuclease. Immunoreactivity of pancreas-specific antigen was sensitive to proteolytic enzymes, perchloric acid and high temperature (70 degrees C, 10 min); but insensitive to neuraminidase or beta-glucosidase. Immunohistochemical staining revealed that pancreas-specific antigen was located in acinar cells of human pancreas. In addition, a higher concentration of pancreas-specific antigen was detected in pancreatic juice than in the saline extract of pancreas. This newly identified pancreas-specific antigen, therefore, may be a useful marker protein in physiological studies of pancreas and pancreatic secretion.

The role of prior knowledge and elaboration in text comprehension and memory: a comparison of self-generated elaboration and text-provided elaboration.

Four experiments investigated the interactive effect of text-provided elaborations and prior knowledge on text comprehension and memory. Subjects read 28 episodes, half of which were associated with well-known people and the other half with unknown people. In Experiment 1, text-provided elaborations enhanced recall only when the reader did not have a high level of prior knowledge. The findings from Experiment 1 were hypothesized to be the result of readers generating relevant elaborations during text comprehension. Experiment 2 supported this hypothesis by providing evidence of self-generated elaborations. The results from Experiments 3 and 4 extended these findings by showing that readers with high prior knowledge automatically generate causally relevant elaborations when the sentences have a low relation.

Memory transfer process: an extension of the Sternberg paradigm.

Research conducted by Sternberg (1969) has suggested that there is a transfer process, a cognitive process for activating stored knowledge. His research and that of those who have replicated and extended it have concentrated on distinct declarative lists of information. In two experiments, we have sought to extend this area by studying the cognitive processes for activating an action plan, a sequence of mental operations executed to achieve a goal. Results showed that (a) longer action plans took more time to activate, and (b) when the activation process was interrupted before completion, the process had to be restarted from its beginning after the interruption had been dealt with. These results were interpreted as evidence for a time-consuming transfer process in which the elements of action plans are transferred from long-term store to short-term store. These results are compared with other current data on the transfer process, and implications for the use of controlled knowledge in cognitive tasks are discussed.

A generalized model for the prediction of lead body burdens.

A compartmental model of a typical 70-kg male for lead intake, distribution, and transport has been developed based on previous pharmacokinetic models and experimental results for lead in the human body. A set of first-order, linear ordinary differential equations with constant coefficients is solved to predict lead levels in blood, bone, and other compartments as a function of time resulting from inputs from air and/or ingestion. The model has been shown to be in excellent agreement with the measurements of blood lead for a controlled study by M. B. Rabinowitz et al. (1976, J. Clin. Invest., 58, 260-270). Favorable agreement was also found with blood and urine results reported by T. B. Griffin et al. (1975, "Lead," pp. 221-240) providing that an allowance was made for an unmeasured input of lead, originating from smoking, snacks, etc. The predictions of the newly formulated model are compared with those of the established Bernard model (S. F. Bernard, 1977, Health Phys., 32, 44-46). Predictions of blood lead concentration for short periods (on the order of months) are fitted better by the new model, while both models predict similar behavior over the longer term (on the order of 5 years and greater).

Purification of a human pancreas-specific antigen.

Using an antibody reagent, an organ-specific antigen has been purified from human pancreas. Purification was achieved using ammonium sulfate and ethanol precipitations, concanavalin A-Sepharose affinity chromatography, anion-exchange chromatography and gel filtration. The antigen had a molecular weight of approximately 68,000 as determined by gel filtration, was not dissociated into subunits by sodium dodecyl sulfate, and did not bind to immobilized concanavalin A. The biological significance of this antigen in pancreatic cancer is being studied.