FMRF-amide immunoreactivity in the mammalian gastroenteropancreatic neuroendocrine system

The presence of FMRF-amide, a cardioactive tetrapeptide, was studied by immunocytochemistry in human and rat gastric antrum and pancreas, and in the ovine, bovine, canine and rabbit pancreas. In human and rat gastric antrum, numerous cells contained FMRF-amide immunoreactive material. By staining of serial sections and by double staining, colocalization of immunoreactivity for gastrin and FMRF-amide was observed in part of the gastrin cells. In the pancreas of these and the other species, immunoreactivity for FMRF-amide was located both in acinar and islet endocrine cells. Colocalization of FMRF-amide and pancreatic polypeptide was found in a proportion of pancreatic polypeptide cells in the pancreas. FMRF-amide immunoreactivity never colocalized with the other neurohormonal peptides which occur in the gastric antrum and the pancreas. Our observations show that neuroendocrine cells occur in the gastric antrum and pancreas which are exclusively immunoreactive or gastrin and for pancreatic polypeptide respectively. In addition cells occur which show immunoreactivity for FMRF-amide as well as for gastrin in the gastric antrum and with antiserum to FMRF-amide as well as for pancreatic polypeptide in the pancreas. It is concluded that FMRF-amide antibodies probably recognize a substance in G and PP cells which is not identical but may be structurally related to gastrin and pancreatic polypeptide.     

Immunocytochemical localization of Na+, K+-ATPase in the rat kidney

Summary To determine if rat kidney Na⁺, K⁺-ATPase can be localized by immunoperoxidase staining after fixation and embedding, we prepared rabbit antiserum to purified lamb kidney medulla Na⁺, K⁺-ATPase. When sodium dodecylsulfate polyacrylamide electrophoretic gels of purified lamb kidney Na⁺, K⁺-ATPase and rat kidney microsomes were treated with antiserum (1200), followed by [¹²⁵I]-Protein A and autoradiography, the rat kidney microsomes showed a prominent radioactive band coincident with the -subunit of the purified lamb kidney enzyme and a fainter radioactive band which corresponded to the -subunit. When the Na⁺, K⁺-ATPase antiserum was used for immunoperoxidase staining of paraffin and plastic sections of rat kidney fixed with Bouin's, glutaraldehyde, or paraformaldehyde, intense immunoreactive staining was present in the distal convoluted tubules, subcapsular collecting tubules, thick ascending limb of the loops of Henle, and papillary collecting ducts. Proximal convoluted tubules stained faintly, and the thin portions of the loops of Henle, straight descending portions of proximal tubules, and outer medullary collecting ducts did not stain. Staining was confined to basolateral surfaces of tubular epithelial cells. No staining was obtained with preimmune serum or primary antiserum absorbed with purified lamb kidney Na⁺, K⁺-ATPase, or with osmium tetroxide postfixation. We conclude that the basolateral membranes of the distal convoluted tubules and ascending thick limb of the loops of Henle are the major sites of immunoreactive Na⁺, K⁺-ATPase concentration in the rat kidney.Supported by Grant AM 17047 from NIH and by the Veterans Administration     

FMRF-amide immunoreactivity in the mammalian gastroenteropancreatic neuroendocrine system

Summary The presence of FMRF-amide, a cardioactiv tetrapeptide, was studied by immunocytochemistry in human and rat gastric antrum and pancreas, and in the ovine, bovine, canine and rabbit pancreas. In human and rat gastric antrum, numerous cells contained FMRF-amide immunoreactive material. By staining of serial sections and by double staining, colocalization of immunoreactivity for gastrin and FMRF-amide was observed in part of the gastrin cells. In the pancreas of these and the other species, immunoreactivity for FMRF-amide was located both in acinar and islet endocrine cells. Colocalization of FMRF-amide and pancreatic polypeptide was found in a proportion of pancreatic polypeptide cells in the pancreas. FMRF-amide immunoreactivity never colocalized with the other neurohormonal peptides which occur in the gastric antrum and the pancreas.Our observations show that neuroendocrine cells occur in the gastric antrum and pancreas which are exclusively immunoreactive or gastrin and for pancreatic polypeptide respectively. In addition cells occur which show immunoreactivity for FMRF-amide as well as for gastrin in the gastric antrum and with antiserum to FMRF-amide as well as for pancreatic polypeptide in the pancreas. It is concluded that FMRF-amide antibodies probably recognize a substance in G and PP cells which is not identical but may be structurally related to gastrin and pancreatic polypeptide.In honour of Prof. P. van Duijn     

Characterization and Location of Met5-Enkephalin-Arg6-Phe7 Stored in Various Rat Brain Regions

A specific antiserum against met⁵-enkepha-lin-arg⁶-phe⁷ was raised and used to study the distribution and characterization of met⁵-enkephalin-arg⁶-phe⁷-like immunoreactive material in rat brains by radioimmunoassay and immunohistochemical procedures. The antiserum appears to be directed to the COOH-terminus of the peptide, as it fails to cross-react with met⁵-enkeph-alin, met³-enkephalin-arg⁶, met⁵-enkephalin-arg⁶-arg⁷, met⁶-enkephalin-lys⁶, and leu-enkephalin. However, it cross-reacts with phe-met-arg-phe by about 10% and with phe-met-arg-phe-NH₂ to an insignificant degree. The highest content of met⁵-enkephalin-arg⁶-phe⁷ was found in the striatum, which contains a dense network of immunoreactive varicose fibers and terminals, as well as immunoreac tive cell bodies. The met⁵-enkephalin-arg⁶-phe⁷ in striatum can be released in a Ca²⁺-dependent manner by a depolarizing concentration of KC1, raising the possibility of a neu-roregulatory role for met⁵-enkephalin-arg⁶-phe⁷. Characterization of the immunoreactive material by gel filtration and high pressure liquid chromatography revealed the presence of multiple forms of immunoreactive material in some brain regions.     

The use of bacitracin as an inhibitor of the degradation of thyrotropin releasing factor and luteinizing hormone releasing factor

Bacitracin was found to be an effective inhibitor of the $\text{in}\text{vitro}$ degradation of both thyrotropin releasing factor¹ (TRF) and luteinizing hormone releasing factor (LRF) by guinea pig hypothalamic and whole brain homegenates and rat hypothalamic homogenates and subcellular fractions. Bacitracin was effective in inhibiting the degradation of TRF and LRF, as determined by radioimmunoassay, where it exhibited no interference with the assays. Kinetic studies of the degradation of exogenous synthetic [³H]-TRF demonstrated non-competitive inhibition by bacitracin with K_i = 1.9 × 10^?5 M, while studies on the degradation of [³H] LRF indicated competitive inhibition with K_i = 1.7 × 10^?5 M. Electrophoretic and amino acid analysis revealed that bacitracin itself was not degraded during the course of the $\text{in}\text{vitro}$ incubation.     

Immunocytochemical localization of Na+, K+-ATPase in the rat kidney

To determine if rat kidney Na+, K+-ATPase can be localized by immunoperoxidase staining after fixation and embedding, we prepared rabbit antiserum to purified lamb kidney medulla Na+, K+-ATPase. When sodium dodecylsulfate polyacrylamide electrophoretic gels of purified lamb kidney Na+, K+-ATPase and rat kidney microsomes were treated with antiserum (1:200), followed by [125I]-Protein A and autoradiography, the rat kidney microsomes showed a prominent radioactive band coincident with the alpha-subunit of the purified lamb kidney enzyme and a fainter radioactive band which corresponded to the beta-subunit. When the Na+, K+-ATPase antiserum was used for immunoperoxidase staining of paraffin and plastic sections of rat kidney fixed with Bouin's, glutaraldehyde, or paraformaldehyde, intense immunoreactive staining was present in the distal convoluted tubules, subcapsular collecting tubules, thick ascending limb of the loops of Henle, and papillary collecting ducts. Proximal convoluted tubules stained faintly, and the thin portions of the loops of Henle, straight descending portions of proximal tubules, and outer medullary collecting ducts did not stain. Staining was confined to basolateral surfaces of tubular epithelial cells. No staining was obtained with preimmune serum or primary antiserum absorbed with purified lamb kidney Na+, K+-ATPase, or with osmium tetroxide postfixation. We conclude that the basolateral membranes of the distal convoluted tubules and ascending thick limb of the loops of Henle are the major sites of immunoreactive Na+, K+-ATPase concentration in the rat kidney.     

Localization of γ-melanocyte stimulating hormone (γMSH) immunoreactivity in rat brain and pituitary

γ-Melanocyte stimulating hormone (γMSH) is a possibly biologically active material discovered in the cryptic N-terminus of the pro-opiocortin precursor by recombinant DNA analysis of bovine pituitary mRNA. Well-characterized antisera to synthetic bovine γ-3MSH (γ₃MSH) were used to localize immunoreactive sites in sections of formaldehyde-fixed rat brain and pituitary by the indirect immunoperoxidase technique. Specificity for staining was established by absorption with the synthetic antigen peptides or their fragments; staining was not blocked by absorption with synthetic replicates of other natural peptides that contain redundant amino acid sequences, with those of γMSH such as corticotropin or β-MSH. The general patterns of staining within adenohypophysis, intermediate lobe, and central nervous system closely followed the previously described patterns of β-endorphin immunoreactivity. Corticotrophs, all intermediate lobe cells and neuronal perikarya in the ventro-basal hypothalamus exhibited immunoreactivity for γ₃MSH as they do for β-endorphin. Furthermore, the general distribution of immunoreactive nerve fibers and terminals within the diencephalon and pons was quite similar to endorphin immunoreactivity patterns as well. In series of alternating sections, prepared for either γ₃MSH β-endorphin immunoreactivity, the same specific terminal fields were found to exhibit very similarly shaped varicose axons and probable terminal bouton configurations. However, the density of the innervation by fibers exhibiting immunoreactivity for the two peptides varied among the common target areas. Furthermore, the perikarya exhibiting γ₃MSH immunoreactivity were fewer in number, smaller in size, and more medially clustered than those exhibiting immunoreactivity for β-endorphin. These results demonstrate that γ₃MSH also occurs in rat brain and in pituitary cells which were already known to contain endorphin immunoreactivity. However, γ₃MSH-immunoreactive neurones may not be coexistent with all endorphin-immunoreactive neurons, and these cells project with varying intensity to common target fields. Such observations are in agreement with the proposal of different processing of a common precursor in different cells.     

Immunohistochemical detection of human epidermal growth factor in submandibular glands and their tumors using a polyclonal antiserum and a monoclonal antibody

Summary We applied immunohistochemical procedures to detect hEGF in salivary glands and pleomorphic adenomas of salivary-gland origin using a polyclonal hEGF antiserum and a monoclonal antibody against hEGF synthesized by applying the synthetic gene technique using Escherichia coli. In normal salivary glands, hEGF was mainly localized in the ductal system (i.e., intercalated, striated, and excretory ducts). The staining intensity and intracellular localization exhibited some variation depending on the fixative used. When a polyclonal hEGF antiserum was used for immunostaining, slight background staining was observed in sections prepared using the fixatives tested. Therefore, precise localization of hEGF was obtainable only in formalin-fixed sections using the monoclonal antibody against hEGF. In pleomorphic adenomas, positive hEGF staining was seen on the luminal side of tumors and in cells of ductal origin; no reactivity was present on the outer side of tumors or in cells of myoepithelial origin. Occasionally, long, spindleshaped tumor cells and chondroidally changed tumor cells also exhibited positive staining for hEGF.     

Radioimmunoassay for Lys8, Asn9, neurotensin 8–13: Tissue and subcellular distribution of immunoreactivity in chickens

A sensitive and specific radioimmunoassay (RIA) for Lys⁸, Asn⁹, neurotensin 8–13 (LANT-6) has been developed which utilizes ¹²⁵I-labeled LANT-6 and rabbit antisera raised towards conjugates of synthetic LANT-6 and bovine thyroglobulin. The antiserum described (TG-22) allows the detection of ca 100 fmol of LANT-6 and crossreacts <0.01% with chicken or bovine NT. Dose-response relationships for the native (chicken) and synthetic peptides were indistinguishable. Using this assay the distribution of immunoreactive LANT-6 (iLANT-6) through various tissues of the chicken was studied and compared with that of chicken NT (iNT) determined by RIA. Both iNT and iLANT-6 were found primarily in the brain and gastrointestinal tract, however, their regional distributions were found to differ. Subcellular distribution studies in homogenates of chicken brain indicated that obth iNT and iLANT-6 were associated with synaptosome-like and vesicle-like particles. In homogenates of small intestine, pancreas and colon iNT and iLANT-6 appeared to be within osmotically sensitive, sedimentable particles. Analyses using high pressure liquid chromatography established that chicken iLANT-6 co-eluted with the synthetic peptide and that similar substances were present in extracts of rat brain and intestine. These results are consistent with “messenger” roles for these peptides.     

Presence of galanin in human pancreatic nerves and inhibition of insulin secretion from isolated human islets

Summary In several animal species, galanin occurs in pancreatic nerves and inhibits insulin secretion. However, the presence and action of galanin in the human pancreas have not been established. Therefore, we examined the presence and nature of human pancreatic galanin-like immunoreactive material (GLIR) and the effects of galanin on glucose-stimulated insulin secretion from isolated human islets. Immunofluorescent staining of human pancreas revealed GLIR in fine varicose fibers in both islets and exocrine parenchyma. Furthermore, acid extracts of pancreas (n=3) and isolated islets (n=3) contained 0.17±0.06 and 0.23±0.11 pmol GLIR/mg protein. Human pancreatic GLIR coeluted with synthetic porcine galanin from Sephadex G-50. Moreover, synthetic porcine galanin inhibited glucose-stimulated insulin secretion from collagenase-isolated human islets at dose rates >10^-8 M. Thus, (1) human pancreas is innervated by galanin-containing nerves, (2) human pancreatic GLIR is of similar size as synthetic porcine galanin, and (3) porcine galanin inhibits glucose-stimulated insulin secretion from human islets. Therefore, galanin could be an important local regulator of insulin secretion in man.     

Species differences in the immunoreactive patterns of calcitonin gene-related peptide in the pancreas

Summary In the pancreas, calcitonin gene-related peptide (CGRP) immunoreactivity has been described in nerve fibers and in distinct types of islet cells. This unique, apparently species-specific cell-type expression prompted the present investigation to clarify further the pattern of CGRP immunoreactivity in different mammalian species (i.e., different strains of rats, mice, guinea pigs, rabbits, cats, dogs, pigs, and humans) commonly used for functional and anatomical studies of the pancreas by means of immunohistochemistry using three different CGRP antibodies. In each species, CGRP-immunoreactive neurites innervate the exocrine and endocrine compartments, the vasculature, and the intrapancreatic ganglia, where they form dense networks encircling unstained cell bodies. The only exception is the pig pancreas, where the islets appear to be devoid of immunoreactive fibers. The overall density of immunoreactive pancreatic axons in different species is as follows: rat, mouse, and rabbit>guinea pigpig and cat> >dog and human. CGRP-immunoreactive endocrine cells appear to be restricted to the rat pancreas, where they form a subpopulation of somatostatin-containing D cells. In contrast, in mouse, guinea pig, cat, dog, and human pancreas, a homogeneous staining of the core of the islets, where insulin-producing B cells are located, was visualized in sections incubated with the rabbit CGRP antiserum at 4°C, but not at 37°C (an incubation temperature that does not affect the islet cell staining in the rat nor the fiber labeling in any species). Furthermore, the staining of islet B cells was not reproductible with all the CGRP antibodies used, all of which comparably stain nerve fibers in each species, and islet D cells in the rat. Immunoreactive islet cells were not visualized in pig and rabbit pancreas. These results are consistent with the hypothesis that the expression of CGRP in nerve fibers is a common feature of mammalian pancreas, whereas its expression in endocrine cells appears to be restricted to the D cells of the rat pancreas.     

Tachykinin-related neuropeptide in the crayfish olfactory midbrain

Immunoreactivity indicative of tachykinin-related peptide (TRP) was detected in the olfactory midbrain of the crayfish Pacifastacus leniusculus when using an antiserum to the insect neuropeptide locustatachykinin I (LomTK-I). A monoclonal antibody to the mammalian tachykinin substance P was shown in double-labeling experiments to label structures in the crayfish brain identical to those labeled with the LomTK antiserum. Within the midbrain LomTK-like immunoreactive (LomTK-LI) material was observed in a limited population of neuronal somata and their varicose processes. A single pair of large interneurons gave rise to numerous varicose LomTK-LI processes innervating a cluster of cell bodies (cluster 10) as well as the olfactory neuropils. The latter neuropil was also innervated by a population of LomTK-LI globuli cells with cell bodies in cluster 9. Radioimmunoassay (RIA), utilizing the LomTK antiserum, and reverse-phase high-performance liquid chromatography (HPLC) were used to partially characterize the immunoreactive material in extract of the portion of the midbrain that houses the olfactory (OL) and accessory (AL) lobes and cell clusters 9 and 10 on the one hand, and in extract of the remaining parts of the brain on the other. Approximately the same amounts of LomTK-LI material were observed for the two extracts. RIA showed that the immunoreactive material of both extracts diluted roughly in parallel to synthetic LomTK-I and HPLC analysis of the extracts revealed immunoreactive material in both tissues which eluted with retention times in the range of synthetic LomTK-I and LomTK-II. These results suggest that TRPs similar to LomTKs are present in the olfactory midbrain of Pacifastacus. The distribution of immunolabeled neuronal structures suggests that in the crayfish, peptide(s) closely related to insect TRPs may act as a neuroactive substance released from nerve fibers in olfactory neuropil areas and at certain neuronal cell bodies.     

A radioimmunoassay for gamma-melanocyte stimulating hormone

A specific radioimmunoassay for γ-melanocyte stimulating hormone-like peptides has been developed. An antiserum raised in rabbit to synthetic bovine γ₃-MSH, one of the possible γ-MSH peptides, specifically recognizes the portion between His⁵ and Arg¹⁴ of γ₃-MSH without significant cross-reaction with other synthetic γ-MSH-like peptides, α-, β-MSH, adrenocorticotropin, and β-endorphin. The usable range of this RIA is 10 pg to 600 pg of synthetic γ₃-MSH. Three immunoreactive γ-MSH peaks were thus found in gel permeation chromatography of the whole bovine pituitary extract.     

LRF producing cells of the mammalian hypothalamus

Summary The demonstration of perikarya of mediocellular neurones producing LRF, using indirect immunofluorescence on slides and anti synthetic LRF antibodies, requires both their activation and the inhibition of their axoplasmic transport. This fact suggests that LRF is present in an immunoreactive form, but usually in very low concentrations. Perikarya of neurons producing LRF are found principally in the preoptic and septal areas of the rat and decrease caudally, particularly beyond the retrochiasmatic area. Most of the axons coming from these perikarya are incorporated in the hypothalamoinfundibular tract and terminate around the capillaries of the primary portal plexus, particularly those of interealar plexus. Other axons (or axon collaterals) may be found in various areas (suprachiasmal crista, epithalamus, amygdala, mesencephalon) and form circuits recalling the “extrahypophyseal pathways” described for the magnocellular Gomori-positive neurons of the SON and PVN. These axons are probably concerned in intersegmental regulations involving “neurosecretory synapses”, particularly of the axosomatic type. The placement of stereotaxic lesions was used to determine the topography and direction of axoplasmic flow of the axons transporting LRF. The infundibular immunoreactive material, already discernible at the end of gestation in the foetus, shows considerable variations between birth and puberty, during the estrous cycle and under various other physiological or experimental conditions. The observations made under various experimental or physiological conditions suggest that, in the guinea-pig in particular, the greater part of the hypothalamic immunoreactive material is concentrated in the infundibular area. This area of accumulation is comparable to the distal neurohypophysis of the Gomori-positive neurosecretory system coming from the SON and PVN. This work was financed by the D.G.R.S.T. Contract No. 72-7-0375.     

Localization of the prostaglandin F2 alpha receptor in rat tissues

The localization of the prostaglandin F_2α (FP) receptor was examined in rat tissues by immunohistochemistry and in situ hybridization. Immunohistochemistry on paraffin sections was performed with a rabbit polyclonal antiserum raised against a synthetic peptide derived from the rat FP receptor sequence. In situ hybridization on cryosections was done with ³⁵S-labelled rat FP receptor antisense and sense riboprobes. The most intense FP receptor-like immunoreactivity was observed in granulosa luteal cells, muscle and epithelial cells, e.g. cardiac, skeletal and smooth muscle, and hepatocytes. Weaker immunoreactivity was found in connective tissue fibroblasts. In the eye, intense immunostaining was associated with the corneal and conjunctival epithelium and moderate staining with the ciliary body, retina, iris and connective tissues. In situ hybridization generally confirmed the results. The riboprobe hybridized weakly with the heart, skeletal muscle, uterus, liver, lung and corpus luteum. Thus, the prostaglandin FP receptor was found to be widely distributed in rat tissues.     

Identification, characterization and localization of thyrotropin-releasing hormone precursor peptides in perinatal rat pancreas

The sequence of rat hypothalamic prepro TRH, deduced from its complementary DNA, contains five TRH progenitor sequences and six cryptic sequences separated by paired basic amino acid residues. We have utilised antisera against two synthetic peptides corresponding to sequences within proTRH, [Tyr53] preproTRH (53-74), part of the amino terminal leader sequence of proTRH and [Cys 74,83] preproTRH-(75-82), representing a TRH progenitor sequence flanked by cysteine residues (pCC10) in radioimmunoassays (RIA) to identify and chromatographically characterize proTRH derived peptides in extracts of rat perinatal pancreas and to localize these peptides immunohistochemically. Two forms of immunoreactive pYT22 (ipYT22) were observed, similar in size to ipYT22 seen in extracts of adult rat brain. By RIA immunoreactive pCC10 was detectable in neonatal but not fetal pancreas. However, immunohistochemical double staining of both fetal and neonatal rat pancreas colocalized both ipYT22 and ipCC10 with immunoreactive insulin in the B-cell of the developing Islets of Langerhans. These findings indicate that the B-cell of the perinatal pancreas synthesizes TRH from a prohormone encoded by a mRNA similar to that present in adult rat hypothalamus.     

NH2-terminal dodecapeptide of porcine big gastrin: Revised sequence and confirmation of structure by immunochemical analysis

A revised sequence for the NH₂-terminal dodecapeptide of porcine big gastrin is described which differs from that originally reported in the inversion of His⁷ and Pro⁹ for Pro⁷ and His⁹. The immunochemical properties of a range of synthetic peptide fragments and analogs of the original and revised sequences of porcine big gastrin were examined with an antiserum raised to the natural porcine peptide. The pattern of immunoreactivity of these peptides indicates that the antiserum has specificity for the 4–9 region of big gastrin. The dodecapeptide with the revised sequence had full immunoreactive potency relative to natural porcine big gastrin, whereas the dodecapeptide with the original sequence had about 1000-fold lower immunoreactivity. It is proposed that the synthetic peptide with the revised, but not the original, sequence is compatible with the structure of big gastrin.     

Cell-free processing of the secreted globins of Chironomus thummi (Diptera)

Poly(A)⁺ RNA from fourth instar larvae of the insect Chironomus thummi was efficiently translated in vitro. Up to 10% of the total cell-free translation products was immunoprecipitable with antiserum specific for chironomid haemoglobins. Six immunoreactive preglobin products were resolved on SDS-17.5% polyacrylamide gels, and were cotranslationally processed by dog pancreas microsomes to 2 major bands that comigrated with authentic secreted globins.     

Localization and characterization of the N-terminal fragment of atrial natriuretic factor (ANF) precursor in the frog heart

The localization of the N-terminal fragment of the atrial natriuretic factor (ANF) precursor in the heart of the frog Rana ridibunda was examined by the indirect immunofluorescence and the immunogold techniques using an antiserum directed against synthetic rat ANF (Asp11-Ala37). At the optic level, positive material was found in most atrial myocytes. Staining of consecutive sections of frog heart with antibodies against N-terminal and C-terminal regions of the proANF molecule showed that both peptides are contained in the same cardiocytes. In the rat atrium, antibodies against the N-terminal ANF region induced a more intense labeling than in the frog atrium. Electron microscopic studies indicated that all secretory granules present in frog atrial cardiocytes contain N-terminal ANF-like immunoreactive material. The positive material localized in frog atrium was characterized by gel filtration and radioimmunological detection. Serial dilutions of frog atrial extracts exhibited displacement curves which were parallel to that obtained with synthetic human ANF (Asn1-Asp30). Sephadex G-50 gel chromatography of the immunoreactive material showed that the N-terminal ANF-like immunoreactivity eluted in a single peak corresponding to high molecular weight material. These results indicate that the N-terminal fragment of frog proANF is immunologically and biochemically related to the homologous mammalian peptide.     

Rapid purification of a prekallikrein from rat pancreas

A prekallikrein from rat pancreas was purified 1500-fold with an overall yield of 20% using a rapid, simple procedure. DEAE-Sephadex A-50 chromatography permitted the separation of two prekallikrein components present in rat pancreatic homogenates; the major fraction was further purified by Sephadex G-100 gel filtration and immunoadsorption chromatography. The zymogen is a single-chain molecule with pI 4·35. Apparent M_r values of 38,000 and 37,000 were determined by sodium dodecyl sulfate-polyacrylamide gradient electrophoresis and gel filtration, respectively.