Anti-cytokine autoantibodies preceding onset of autoimmune polyendocrine syndrome type I features in early childhood.

PURPOSE: Almost all patients with autoimmune polyendocrine syndrome (APS)-I have high titer neutralizing autoantibodies to type I interferons (IFN), especially IFN-ω and IFN-α2, whatever their clinical features and onset-ages. About 90 % also have antibodies to interleukin (IL)-17A, IL-17F and/or IL-22; they correlate with the chronic mucocutaneous candidiasis (CMC) that affects ~90 % of patients. Our aim was to explore how early the manifestations and endocrine and cytokine autoantibodies appear in young APS-I patients. That may hold clues to very early events in the autoimmunization process in these patients. METHODS: Clinical investigations and autoantibody measurements in 13 APS-I patients sampled before age 7 years, and 3 pre-symptomatic siblings with AIRE-mutations in both alleles. RESULTS: Antibody titers were already high against IFN-α2 and IFN-ω at age 6 months in one sibling-8 months before onset of APS-I-and also against IL-22 at 7 months in another (still unaffected at age 5 years). In 12 of the 13 APS-I patients, antibody levels were high against IFN-ω and/or IL-22 when first tested, but only modestly positive against IFN-ω in one patient who had only hypo-parathyroidism. Endocrine organ-specific antibodies were present at age 6 months in one sibling, and as early as 36 and 48 months in two of the six informative subjects. CONCLUSION: This is the first study to collate the onset of clinical features, cytokine and endocrine autoantibodies in APS-I infants and siblings. The highly restricted early autoantibody responses and clinical features they show are not easily explained by mere loss of broad-specific self-tolerance inducing mechanisms, but hint at some more sharply focused early event(s) in autoimmunization.

An autosomal locus causing autoimmune disease: autoimmune polyglandular disease type I assigned to chromosome 21.

Autoimmune polyglandular disease type I (APECED) is an autosomal recessive autoimmune disease characterized by a variable combination of the failure of the endocrine glands. The pathogenesis of this unique autoimmune disease is unknown; unlike many other autoimmune diseases, APECED does not show association to specific HLA haplotypes. Unravelling the APECED locus will identify a novel gene outside the HLA loci influencing the outcome of autoimmune diseases. We have assigned the disease locus to chromosome 21q22.3 by linkage analyses in 14 Finnish families. Linkage disequilibrium studies have significantly increased the informativeness of the analyses and helped to locate the critical DNA region for the APECED locus to just 500 kilobases, a much more precise definition than linkage analyses alone could achieve.

Gene encoding a new RING-B-box-Coiled-coil protein is mutated in mulibrey nanism.

Mulibrey nanism (for muscle-liver-brain-eye nanism, MUL; MIM 253250) is an autosomal recessive disorder that involves several tissues of mesodermal origin, implying a defect in a highly pleiotropic gene. Characteristic features include severe growth failure of prenatal onset and constrictive pericardium with consequent hepatomegaly. In addition, muscle hypotonia, J-shaped sella turcica, yellowish dots in the ocular fundi, typical dysmorphic features and hypoplasia of various endocrine glands causing hormonal deficiency are common. About 4% of MUL patients develop Wilms' tumour. MUL is enriched in the Finnish population, but is rare elsewhere. We previously assigned MUL to chromosome 17q22-q23 and constructed a physical contig over the critical MUL region. The region has now been further refined by haplotype analysis and new positional candidate genes have been localized. We identified a gene with four independent MUL-associated mutations that all cause a frameshift and predict a truncated protein. MUL is ubiquitously expressed and encodes a new member of the RING-B-box-Coiled-coil (RBCC) family of zinc-finger proteins, whose members are involved in diverse cellular functions such as developmental patterning and oncogenesis.

TSGA10 - A target for autoantibodies in autoimmune polyendocrine syndrome type 1 and systemic lupus erythematosus.

Autoimmune polyendocrine syndrome type 1 (APS1) is a rare monogenic autoimmune disorder caused by mutations in the autoimmune regulator (AIRE) gene. High-titre autoantibodies are a characteristic feature of APS1 and are often associated with particular disease manifestations. Pituitary deficits are reported in approximately 7% of APS1 patients, with immunoreactivity to pituitary tissue frequently described. Using APS1 patient serum to immunoscreen a pituitary cDNA expression library, testis specific, 10 (TSGA10) was isolated. Immunoreactivity against TSGA10 was detected in 5/99 (5.05%) patients with APS1, but also in 5/135 (3.70%) systemic lupus erythematosus (SLE) patients and 1/188 (0.53%) healthy controls. TSGA10 autoantibodies were not detected in the serum from patients with any other autoimmune disease. Autoantibodies against TSGA10 were detectable from a young age in 4/5 positive APS1 patients with autoantibody titres remaining relatively constant over time. Furthermore, real-time PCR confirmed TSGA10 mRNA to be most abundantly expressed in the testis and also showed moderate and low expression levels throughout the entire body. TSGA10 should be considered as an autoantigen in a subset of APS1 patients and also in a minority of SLE patients. No recognizable clinical phenotype could be found to correlate with positive autoantibody reactivity.

Risk of autoimmune diabetes in APECED: association with short alleles of the 5'insulin VNTR.

Autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) is a rare autoimmune disease causing a wide spectrum of autoimmune dysfunction potentially including diabetes of an autoimmune etiology. We have previously described a pair of discordant APECED siblings and pointed to a possible role of 5'insulin variable number of tandem repeats (VNTR) locus IDDM2 in the appearance of diabetes within this disease. In vitro studies have previously suggested that class I VNTR alleles were associated with decreased fetal thymic insulin expression. We genotyped the 5'INS VNTR locus and several flanking 11p15.5 markers in 50 Finnish APECED subjects and explored the possible contribution of IDDM2 in the development of diabetes. The shorter 5'INS VNTR class I alleles (<35 repeats) were more prevalent in the diabetic Finnish APECED subjects than in non-diabetic APECED subjects. Logistic regression analysis revealed that having 1 short (<35) VNTR allele did not increase the risk of developing diabetes (95% CI 0.6-27.0), whereas having 2 short alleles conferred a 43.5-fold increased risk (95% CI 3.0-634.6). We conclude that short 5'INS VNTR class I alleles play a role in susceptibility to autoimmune diabetes in the context of APECED.

Clinical manifestations and management of patients with autoimmune polyendocrine syndrome type I.

Autoimmune polyendocrine syndrome type I (APS-I) is a monogenic model disease of autoimmunity. Its hallmarks are chronic mucocutaneous candidosis, hypoparathyroidism and adrenal insufficiency, but many other autoimmune disease components occur less frequently. The first components usually appear in childhood, but may be delayed to adolescence or early adult life. There is enormous variation in presentation and phenotype, which makes the diagnosis difficult. Antibodies against interferon-omega and -alpha have recently been shown to be sensitive and relatively specific markers for APS-I, and mutational analysis of the autoimmune regulator gene gives the diagnosis in >95% of cases. The treatment and follow-up of patients is demanding and requires the collaboration of specialists of several fields. However, the literature is especially sparse regarding information on treatment and follow-up; hence, we present here a comprehensive overview on clinical characteristics, treatment and follow-up based on personal experience and published studies.

Autoimmune polyendocrinopathy-candidosis-ectodermal dystrophy (APECED)--a diagnostic and therapeutic challenge.

Autoimmune polyendocrinopathy - candidosis - ectodermal dystrophy (APECED), also known as autoimmune polyendocrine/polyglandular syndrome type 1 (APS1), is a rare disease caused by mutations in the autoimmune regulator (AIRE) gene pair resulting in absence of active AIRE protein, which is essential for both central and peripheral self-tolerance. The phenotype is widely variable. Apart from the classical triad of mucocutaneous candidosis, hypoparathyroidism and adrenal failure, several other components, some of which are potentially life-threatening, may develop. Due to the unpredictable clinical course, the patients need regular follow-up by a clinician familiar with the disease. Diagnosis is often possible by clinical diagnostic criteria, but in many cases the early clinical picture does not bring it to mind. A novel tool, search for autoantibodies against interferon-omega, enables proof or exclusion of APECED with more certainty than gene analysis. It is highly specific and sensitive for APECED if thymoma and myasthenia gravis are excluded.

Renal handling of diamino acids in lysinuric protein intolerance.

Lysinuric protein intolerance (LPI) is a rare recessively inherited disease in which one of the fundamental physiological defects is in the mechanism by which diamino acids are transported by the kidney. The purpose of the present studies was to examine that mechanism in four controls and seven patients with LPI. Two types of studies were conducted. In the first set, the renal handling of l-arginine and l-ornithine was evaluated by gradually increasing the plasma concentration of each of these amino acids by constant infusion techniques. In the second set of studies, the possible existence of competitive inhibition between l-arginine, l-ornithine, and l-lysine was examined. In the control subjects, there was almost complete reabsorption of arginine and ornithine, with increases in their filtered loads to 50-100 times normal. With further increases in the filtered loads of these amino acids, there was a gradual decrease in their fractional reabsorption. Mutual competitive inhibition was suggested by the observation that an increase in the filtered load of one diamino acid was associated with a decrease in the reabsorption of the other two. In LPI, the fasting plasma diamino acid concentrations were significantly lower than in the controls. With low filtered loads, the fractional reabsorption of the diamino acids was clearly below normal. This defect diminished with higher loads. A stepwise increase in the plasma concentration of one diamino acid resulted in a biphasic response. Initially, net tubular secretion of the other diamino acids was noted, but later was followed by return to net absorption. When two diamino acids were infused simultaneously, net absorption of both took place, though less efficiently than in the controls. We conclude that the renal reabsorption mechanism is defective in patients with LPI. With low normal filtered loads, there is increased fractional excretion of all three diamino acids resulting in low serum concentrations of these compounds. However, at higher artificially elevated concentrations of diamino acids, the capacity of the renal transport system in these patients appears normal.

Colonic electrolyte transport in health and in congenital chloride diarrhea.

Congenital chloride diarrhea (CCD) is a rare autosomal recessive disorder, characterized by watery stools with C1- concentration around 150 meq/liter. We have perfused the colon of three patients and their three healthy siblings with different salt solutions containing 36C1- to determine the nature of the colonic defect in CCD. In the controls, net absorption of Na+ and C1- occurred against steep concentration gradients. The influx (lumen-to-plasms flux) of C1- was twice the effux. Omission of HCO3- from the perfusate caused a clear decrease in C1- efflux which suggests a coupling of C1- effux to HCO3- influx. In CCD, net Na+ absorption occurred normally when HCO3- was present in the lumen. However, Na+ absorption was always impaired when the luminal contents were acid, a situation that prevails in CCD. Net K+ secretion was clearly increased. Both influx and efflux of C1- were practically absent. Only slight net secretion occurred along a steep gradient. Net appearance of HCO3- was not observed, in contrast to controls. These findings and earlier studies of ileal function in CCD are best explained by a defect in the C1-/HCO3- exchange mechanism, which operates in both directions in the normal ileum and colon.

Lysinuric protein intolerance. Basolateral transport defect in renal tubuli.

In patients with an autosomal recessive diamino acid transport disorder, lysinuric protein intolerance (LPI), we measured plasma and urinary amino acids basally, and during intravenous infusion of citrulline at two rates. Compared with controls, the patients' plasma citrulline concentrations rose similarly, but urinary citrulline excretion increased excessively. Their plasma arginine and ornithine levels rose subnormally, but massive argininuria and moderate ornithinuria appeared. The excretion rates of the third diamino acid lysine and other amino acids remained practically unaltered, thus excluding mutual competition as the cause for the increases. The results suggest that (a) in the normal kidney reabsorption involves partial conversion of citrulline to arginine and ornithine (metabolic run-out), (b) in LPI, the diamino acid transport defect is located at the basolateral cell membrane of the renal tubules; this inhibits the efflux of arginine and ornithine, increasing their cellular concentration, which in turn inhibits the metabolic disposal of citrulline, and causes leakage of arginine, ornithine, and citrulline into the tubular lumen.

Lysine fluxes across the jejunal epithelium in lysinuric protein intolerance.

Lysinuric protein intolerance (LPI) is one of a group of genetic diseases in which intestinal absorption of the diamino acids lysine, arginine, and ornithine is impaired. In LPI, the clinical symptoms are more severe than in the kindred disorders. The mechanism of lysine absorption was, therefore, investigated in vitro on peroral jejunal biopsy specimens in seven patients with LPI and 27 controls. The lysine concentration ratio between cell compartment and medium was significantly higher in the LPI group (mean+/-SEM, 7.17+/-0.60) than in the controls (5.44+/-0.51). This was also true for the intracellular Na concentration (LPI, 73.6+/-10.8 mM; controls 42.3+/-3.7 mM). The rate of unidirectional influx of lysine across the luminal membrane was Na dependent and was the same in the two groups. In the absence of an electrochemical gradient, net transepithelial lysine secretion was observed in LPI. This was entirely the result of a 60% reduction of the unidirectional flux from mucosa to serosa. Calculation of unidirectional fluxes revealed the most striking difference at the basolateral membrane, where the flux from cells to serosa was reduced by 62% and the corresponding permeability coefficient reduced by 71%. A progressive reduction in short-circuit current appeared in the epithelia of all four patients with LPI tested after addition of 3 mM lysine. Thus, LPI appears to be the first disease in which a genetically determined transport defect has been demonstrated at the basolateral membrane.

Glutamate decarboxylase-reactive peripheral blood lymphocytes from patients with IDDM express gut-specific homing receptor alpha4beta7-integrin.

Migration of lymphocytes to the pancreas is a prerequisite for insulitis in IDDM. Mucosal vascular addressin (MAdCAM-1), involved in the recirculation of lymphocytes to the gut, has been found in the inflamed islets in NOD mice. In humans, triggers of the gut immune system (e.g., early exposure to cow's milk proteins in infancy, exposure to enteroviral infections) have been associated with IDDM. To study the possible link between the gut immune system and IDDM, we tested the expression of the alpha4beta7-integrin, a homing receptor for MAdCAM-1, on GAD65-reactive lymphocytes. Using immunomagnetic cell sorting, we depleted the lymphocytes with high expression of alpha4beta7-integrin in the peripheral blood mononuclear cell population from IDDM patients and patients with autoimmune polyendocrine disease type 1 (APD-I). The depletion led to a marked decrease (mean 70%) in the cellular response against GAD65 in three of six IDDM patients and in one subject at high risk for IDDM. A decrease of 37% in the GAD response was observed after depletion in the case of one APD-I patient who also had IDDM. Cellular response to tetanus toxoid increased in the majority of patients as well as in three control subjects studied. We demonstrated that a remarkable population of islet cell antigen-reactive lymphocytes express the gut-specific homing receptor, which emphasizes the role of gut immunity in IDDM. The manipulation of the gut immune system is therefore proposed as a tool for modulation of the autoimmunity against pancreatic beta-cells in IDDM.

Mulibrey nanism: clinical features and diagnostic criteria.

Mulibrey nanism (MUL) is an autosomal recessive disease caused by mutations in the TRIM37 gene encoding the peroxisomal TRIM37 protein of unknown function. In this work, we analysed the clinical characteristics of 85 Finnish patients with MUL, most of whom were homozygous for the Finn major mutation of TRIM37. The patients' hospital records from birth to the time of the diagnosis at age 0.02-52 years (median 2.1 years) were retrospectively analysed. All except four of the patients (95%) had a prenatal onset growth failure without postnatal catch up growth. The mean length standard deviation score (SDS) was -3.1 and -4.0 at birth and at diagnosis, respectively. In infancy, feeding difficulties, and respiratory tract infections were the most common problems. Congestive heart failure and pericardial constriction were diagnosed during infancy in 12% and 6% of the patients, respectively. At the time of the diagnosis, characteristic craniofacial features of scaphocephaly, facial triangularity, high and broad forehead, and low nasal bridge were evident in over 90% of the patients. In addition, practically all patients were gracile and had thin extremities. Other findings included a peculiar high-pitched voice (96%), yellowish dots in ocular fundi (79%), cutaneous naevi flammei (65%), hepatomegaly (45%), and fibrous dysplasia of long bones (25%). Mild muscular hypotonicity (68%) was the only neurological abnormality. The clinical features of the Finnish patients with MUL formed a distinct entity. The most consistent findings were growth failure and characteristic craniofacial features. However, organ manifestations varied considerably in early childhood. Based on these findings, we propose new diagnostic criteria for MUL.

Antibodies against hair follicles are associated with alopecia totalis in autoimmune polyendocrine syndrome type I.

In the autosomal recessively inherited autoimmune polyendocrine syndrome type I (APS I) patients have autoantibodies directed against several endocrine and nonendocrine organs. Alopecia areata is present in about one-third of the patients and usually in the more severe forms, alopecia universalis or totalis. Sera from 39 patients with APS I, diluted 1:150, were used in indirect immunofluorescence staining of cryo-sections from normal human scalp. Two hair follicle staining patterns were observed. A cytoplasmic staining of the differentiating matrix, cuticle, and cortex keratinocytes in the anagen hair follicle was seen in five (13%) APS I sera. All these five patients had alopecia totalis, representing 63% of the eight patients with alopecia totalis (p < 0.0001). Furthermore, four (10%) of the APS I sera stained the nuclei of the melanocytes in the hair follicle. Two of these patients had vitiligo. None of 20 healthy control sera stained the keratinocyte cells or the melanocyte nuclei. These data show that many patients with APS I have high-titer autoantibodies directed against the anagen matrix, cuticle, and cortex keratinocytes and a melanocyte nuclear antigen, and also that the hair follicle keratinocyte staining is associated with alopecia, especially alopecia totalis. This study emphasizes the role of the differentiating anagen keratinocytes as an important structure in the autoimmune etiology of alopecia, both in APS I and at least in a subgroup of patients with alopecia areata unrelated to APS I.

Adrenal and steroidal cell antibodies in patients with autoimmune polyglandular disease type I and risk of adrenocortical and ovarian failure.

Thirty-one patients with autoimmune polyglandular disease type I who initially had no adrenocortical and/or ovarian failure were followed for 1.2-12.1 yr (mean, 8.3) by determinations of adrenal (AA) and steroidal cell antibodies (SCA) and functional tests. Adrenocortical failure developed in 13 and ovarian failure in 11 patients. SCA or AA preceded adrenocortical failure in 12 of the 13 patients and were found in 2 of 9 patients (so far) who still have normal adrenal function (P = 0.001). SCA preceded ovarian failure in all 11 patients and were found in 6 of 11 patients who still have normal ovarian function (P = 0.02). The sensitivities/specificities/predictive values were 0.77/0.78/0.90 in all patients for SCA predicting adrenocortical failure, and 0.92/0.89/0.92 for adrenal-binding antibody (which includes all AA and most SCA) in predicting adrenocortical failure. The sensitivities/specificities/predictive values in females who initially had normal adrenocortical and ovarian function were 1.0/0.56/0.50 for SCA in predicting ovarian failure, 0.86/0.83/0.86 for SCA in predicting adrenocortical failure, and 1.0/1.0/1.0 for adrenal-binding antibody in predicting adrenocortical failure. Thus, the appearance of AA or SCA in a male patient without adrenocortical failure or a female patient without adrenocortical or ovarian failure signals a high risk of their development.

Single versus repeated dose human chorionic gonadotropin stimulation in the differential diagnosis of hypogonadotropic hypogonadism.

The responses of serum testosterone (T), 17 alpha-hydroxyprogesterone, and 17 beta-estradiol (E2) to four im injections of hCG (5000 IU/1.7 m2) given on days 0, 4, 7, and 10 were studied in 10 prepubertal and 10 pubertal boys with hypogonadotropic hypogonadism (groups O and P, respectively). Serum was obtained before each injection and on day 14. The results were compared with those of controls, 16 prepubertal boys with incomplete testicular descent and 6 pubertal boys with constitutional delay of puberty. Serum T levels increased significantly in groups O and P to 2.0 and 4.6 nmol/liter, respectively, after the first injection, then progressively to 5.8 and 11.2 nmol/liter. Basal T levels of group O did not differ from those of the controls, but were subnormal for group P (P less than 0.001). Stimulated T levels were subnormal in both groups (P less than 0.01 and P less than 0.001), but repeated doses increased the difference from the control value only in group P. A difference in E2 response between patients and controls appeared in puberty; only the pubertal control boys had substantial increases in E2 (P less than 0.001). Our results show that the optimal protocol for a diagnostic hCG test in prepubertal boys is a single dose of hCG, with determination of T levels 4 days later. In puberty, if the basal T levels are inconclusive, repeated doses of hCG should be given with determination of both T and E2. These findings also suggest that the full inhibitory effect of E2 on T synthesis results from a pubertal maturation process, possibly induced by endogenous gonadotropins, which cannot be induced by two weeks of hCG stimulation in prepubertal boys or those with hypogonadotropic hypogonadism.

Inappropriate secretion of antiduretic hormone, hypertension, and hypoplastic corpus callosum.

The threshold of serum osmolality causing release of vasopressin (antidiuretic hormone) was shifted to an abnormally low level (262 mosmol/kg H2O) in a 14-year-old girl with hypertension and signs of hypoplastic corpus callosum. There was a physiologically meaningful control of vasopressin release in response to water restriction and water load. Plasma vasopressin concentrations (range 1.2--11.9 pg/ml) were of the same magnitude as those of healthy adults, being abnormally high only when related to the hypotonicity of serum observed. Plasma concentrations of angiotensin II were higher than expected from the suppressed levels of plasma renin activity. Blood-pressure response to angiotensin II infusion was increased. Resetting of the osmostat and hypertension may both be explained by lesions of the central nervous system.

The expression of autoimmune polyglandular disease type I appears associated with several HLA-A antigens but not with HLA-DR.

We studied HLA-A, -B, -C, and -DR antigens in 45 patients (from among 34 families), aged 10.2-60 yr, with polyglandular autoimmune disease type I (APG I) and in other family members. HLA-A28 was more frequent in the patients (25%) than in unaffected siblings (16%; P less than 0.05) or in normal Finnish subjects (8.8%; P less than 0.005, corrected P less than 0.2). Compared with the normal subjects, HLA-A28 was more frequent in the patients with hypoparathyroidism (31%; P less than 0.001, corrected P less than 0.04), adrenocortical failure (27%; P less than 0.01), insulin-dependent diabetes mellitus (IDDM; 66%; P less than 0.01), keratopathy (53%; P less than 0.001, corrected P less than 0.04), and alopecia (40%; P less than 0.001, corrected P less than 0.04), but not in the patients with ovarian failure (9%; P = NS). HLA-A28 was more frequent in the patients with hypoparathyroidism (31%) than in APG I patients without it (13%; P less than 0.005, corrected P less than 0.2). It was also more frequent in the patients with IDDM (66%) than in those without it (21%; P less than 0.05). HLA-A3 was more frequent in the patients with ovarian failure (82%) than in APG I patients with normal ovarian function (22%; P less than 0.025) and in normal subjects (45.5%; P less than 0.05). HLA-A9 was less frequent in the patients with ovarian failure (0%) than in those with normal ovarian function (55%; P less than 0.005, corrected P less than 0.2), and it was less frequent (P less than 0.025) in the patients with adrenocortical failure than in those with normal adrenal function. No association was found with any single DR antigen, but of 4 DR-typed IDDM patients, 3 were DR3 or DR4 positive (P = NS). The occurrence of adrenocortical failure, but not hypoparathyroidism, was familial and associated with HLA haploidentity among sets of affected siblings.

Epidermal growth factor in human urine from birth to puberty.

The highest concentrations of epidermal growth factor (EGF) are found in urine, but the physiological role of urinary EGF is unknown. We studied human urinary EGF excretion, by measuring its concentration with a specific homologous RIA, in 265 healthy children from birth until age 16 yr. The absolute concentrations varied widely between individuals. Mean values were approximately 10 ng/ml in 1- to 30-day-old infants; 2.5-fold higher values were found in infants aged 2 to 12 months. During the second year there was a further rise to about 70 ng/ml, and urinary EGF excretion was in the same range in older subjects. The EGF/creatinine concentration ratio was less variable. The mean ratio increased 6-fold from birth to the second year of life. Thereafter, the EGF/creatinine ratio decreased gradually to one-third of the peak level at puberty. No sex difference was found.

Characterization of material with epidermal growth factor immunoreactivity in human serum and platelets.

We investigated the molecular nature of the epidermal growth factor immunoreactivity (ir-EGF) in serum and platelets of normal subjects. In serum, ir-EGF appeared and increased during spontaneous blood coagulation, reaching a plateau in 2 h. The mean plateau measured by time-resolved immunofluorometric assay was 778 pg/mL [130 pmol/L; range, 465-1352 pg/mL (78-225 pmol/L)] for men (n = 66), and 774 pg/mL (129 pmoL/L; range, 521-1114 pg/mL (87-186 pmol/L)] for women (n = 33). Serum samples (n = 9), when examined under nonreducing conditions by high performance liquid chromatography, contained three mol wt components. Their approximate proportions averaged 56 +/- 6% (+/- SD) for a 140K component, 22 +/- 9% for a 67K component, and 22 +/- 6% for a component coeluting with the 6K EGF standard. The molecular size distribution of ir-EGF released from isolated platelets varied with the treatment of the platelets. After stimulation with a Ca2+ ionophore, the proportions of the 140K, 67K, and 6K components were 61 +/- 11%, 20 +/- 8%, and 17 +/- 4% (n = 5), respectively, as in serum. In addition, we found a small amount, (approximately 1%) of a 17K component. When platelets (n = 6) were ruptured by sonication and repeated freeze-thawing, the 67K component formed 53 +/- 7% of the total; the proportions of the 140K, 17K, and 6K components were 14 +/- 7%, 2.1 +/- 0.7%, and 31 +/- 4%, respectively. Under reducing conditions the 17K and 6K components remained intact, but part of the 140K component and all of the 67K component were cleaved to a 35-37K form. After protease digestion of the higher mol wt components, the ir-EGF was exclusively of the 6K form. The 67K and 6K components bound to the EGF receptor, whereas the 140K component bound inconsistently. We conclude that treatment of platelets with Ca2+ ionophore produces ir-EGF components similar to those found in incubated serum but different from those obtained by freeze-thawing. The 67K mol wt component appears to be the main storage form of EGF in platelets; at least two independent mechanisms appear to exist for EGF release.