Starvation actively inhibits splicing of glucose-6-phosphate dehydrogenase mRNA via a bifunctional ESE/ESS element bound by hnRNP K

Regulated expression of glucose-6-phosphate dehydrogenase (G6PD) is due to changes in the rate of pre-mRNA splicing and not changes in its transcription. Starvation alters pre-mRNA splicing by decreasing the rate of intron removal, leading to intron retention and a decrease in the accumulation of mature mRNA. A regulatory element within exon 12 of G6PD pre-mRNA controls splicing efficiency. Starvation caused an increase in the expression of heterogeneous nuclear ribonucleoprotein (hnRNP) K protein and this increase coincided with the increase in the binding of hnRNP K to the regulatory element and a decrease in the expression of G6PD mRNA. HnRNP K bound to two C-rich motifs forming an ESS within exon 12. Overexpression of hnRNP K decreased the splicing and expression of G6PD mRNA, while siRNA-mediated depletion of hnRNP K caused an increase in the splicing and expression of G6PD mRNA. Binding of hnRNP K to the regulatory element was enhanced in vivo by starvation coinciding with a decrease in G6PD mRNA. HnRNP K binding to the C-rich motifs blocked binding of serine-arginine rich, splicing factor 3 (SRSF3), a splicing enhancer. Thus hnRNP K is a nutrient regulated splicing factor responsible for the inhibition of the splicing of G6PD during starvation.

Dietary regulation of expression of glucose-6-phosphate dehydrogenase

The family of enzymes involved in lipogenesis is a model system for understanding how a cell adapts to dietary energy in the form of carbohydrate versus energy in the form of triacylglycerol. Glucose-6-phosphate dehydrogenase (G6PD) is unique in this group of enzymes in that it participates in multiple metabolic pathways: reductive biosynthesis, including lipogenesis; protection from oxidative stress; and cellular growth. G6PD activity is enhanced by dietary carbohydrates and is inhibited by dietary polyunsaturated fats. These changes in G6PD activity are a consequence of changes in the expression of the G6PD gene. Nutrients can regulate the expression of genes at both transcriptional and posttranscriptional steps. Most lipogenic enzymes undergo large changes in the rate of gene transcription in response to dietary changes; however, G6PD is regulated at a step subsequent to transcription. This step is involved in the rate of synthesis of the mature mRNA in the nucleus, specifically regulation of the efficiency of splicing of the nascent G6PD transcript. Understanding the mechanisms by which nutrients alter nuclear posttranscriptional events will help uncover new information on the breadth of mechanisms involved in gene regulation.

Regulation of the processing of glucose-6-phosphate dehydrogenase mRNA by nutritional status

Expression of glucose-6-phosphate dehydrogenase (G6PD) gene during starvation and refeeding is regulated by a posttranscriptional mechanism occurring in the nucleus. The amount of G6PD mRNA at different stages of processing was measured in RNA isolated from the nuclear matrix fraction of mouse liver. This nuclear fraction contains nascent transcripts and RNA undergoing processing. Using a ribonuclease protection assay with probes that cross an exon-intron boundary in the G6PD transcript, the abundance of mRNAs that contain the intron (unspliced) and without the intron (spliced) was measured. Refeeding resulted in 6- and 8-fold increases in abundance of G6PD unspliced and spliced RNA, respectively, in the nuclear matrix fraction. However, the amount of G6PD unspliced RNA was at most 15% of the amount of spliced RNA. During refeeding, G6PD spliced RNA accumulated at a rate significantly greater than unspliced RNA. Further, the amount of partially spliced RNA exceeded the amount of unspliced RNA indicating that the enhanced accumulation occurs early in processing. Starvation and refeeding did not regulate either the rate of polyadenylation or the length of the poly(A) tail. Thus, the G6PD gene is regulated during refeeding by enhanced efficiency of splicing of its RNA, and this processing protects the mRNA from decay, a novel mechanism for nutritional regulation of gene expression.

Polyunsaturated fatty acids inhibit the expression of the glucose-6-phosphate dehydrogenase gene in primary rat hepatocytes by a nuclear posttranscriptional mechanism

Expression of the glucose-6-phosphate dehydrogenase (G6PD) gene is inhibited by the addition of polyunsaturated fatty acids to the medium of primary hepatocytes in culture. To define the regulated step, we measured the abundance of G6PD mRNA both in the nucleus and in total RNA and measured the transcriptional activity of the G6PD gene. Insulin and glucose stimulated a 5- to 7-fold increase in G6PD mRNA in rat hepatocytes. This increase was attenuated by 60% due to the addition of arachidonic acid. These changes in mRNA accumulation occurred in the absence of changes in the rate of transcription. Amounts of precursor mRNA (pre-mRNA) for G6PD in the nucleus changed in parallel with the amount of mature mRNA. The decrease in G6PD pre-mRNA accumulation caused by arachidonic acid was also observed with other long chain polyunsaturated fatty acids but not with monounsaturated fatty acids. In addition, this decrease was not due to a generalized toxicity of the cells due to fatty acid oxidation. These changes in G6PD expression in the primary hepatocytes are qualitatively and quantitatively similar to the changes observed in the intact animal due to dietary carbohydrate and polyunsaturated fat. Regulation of G6PD expression by a nuclear posttranscriptional mechanism represents a novel form of regulation by fatty acids.

Structural characterization and tissue-specific expression of the mouse glucose-6-phosphate dehydrogenase gene

Glucose-6-phosphate dehydrogenase (G6PD) activity differs among tissues and, in liver, with the dietary state of the mouse. Tissue-specific differences in G6PD activity in adipose tissue, liver, kidney, and heart were associated with similar differences in the amount of G6PD mRNA. Regulation of mRNA amount by dietary fat was only observed in liver. In mice fed a low-fat diet, the relative amounts of G6PD mRNA were 3:1:1:0.38, respectively, in the four tissues. Further, the amount of precursor mRNA for G6PD in liver, kidney, and heart reflected the amount of mature mRNA in these tissues, suggesting differing transcriptional activity. Our S1 nuclease and primer-extension analyses indicated that the same transcriptional start site is used in liver, kidney, and adipose tissue, resulting in a common 5' end of the mRNA in these tissues. Thus, differential regulation is not attributable to alternate promoter usage. A DNase hypersensitivity analysis of the 5' end of the G6PD gene identified three hypersensitive sites (HS): HS 1 and HS 2 were present in all tissues, whereas HS 3 was liver specific. Thus, regulation of G6PD expression involves both dietary and tissue-specific signals that appear to act via different mechanisms.

Nutritional regulation of the glucose-6-phosphate dehydrogenase gene is mediated by a nuclear posttranscriptional mechanism

Expression of the glucose-6-phosphate dehydrogenase (G6PD) gene is inhibited by addition of polyunsaturated fat to a high-carbohydrate diet and stimulated by feeding a high-carbohydrate diet to starved mice. The mechanism of this regulation is posttranscriptional. To define the regulated step, we measured the abundance of G6PD mRNA both in the nucleus and in total RNA. Feeding mice a high-fat diet results in a 70% or greater inhibition of nuclear precursor mRNA (pre-mRNA) and mature mRNA abundance. Amounts of both pre-mRNA and mature mRNA for G6PD are stimulated 13-fold or more by refeeding starved mice. Changes in amount of pre-mRNA for G6PD are of a similar magnitude and precede the changes in amount of mature mRNA for G6PD in total RNA. These changes in pre-mRNA abundance occur in the absence of observable changes in the rate of transport of mRNA from the nucleus to the cytoplasm, splicing of the pre-mRNA, or degradation at the 3'-end of the transcript. Despite large changes in pre-mRNA amount in mice fed a low-fat diet relative to mice fed a high-fat diet, the rate of change in the amount of pre-mRNA during the diurnal feeding cycle is not altered. Thus, expression of G6PD is regulated at an early step after transcription of the pre-mRNA. We suggest that pre-mRNA which enters the processing pathway is stable and can be processed and transported to the cytoplasm where it is translated.

Posttranscriptional regulation of glucose-6-phosphate dehydrogenase by dietary polyunsaturated fat

The activity of glucose-6-phosphate dehydrogenase (G6PD) is inhibited by the addition of polyunsaturated fat (PUFA) to a high carbohydrate diet. To define the regulated step, we measured enzyme activity, accumulation of G6PD mRNA, and transcriptional activity of the gene. At steady-state, G6PD activity and mRNA abundance were inhibited by 80% in the livers of mice fed a high-fat diet (6% safflower oil) compared to mice fed a low-fat diet (1% safflower oil). Inhibition of mRNA accumulation was 20% by 4 h and was maximal by 9 h after beginning the high-fat diet. Changes in mRNA accumulation preceded changes in enzyme activity, indicating pretranslational regulation. The rapid kinetics of G6PD mRNA accumulation depended on prior dietary history of the mice. In meal-trained mice, abundance of G6PD mRNA increased by twofold within 4 h of the onset of a low-fat meal and was maximal by 12 h. In contrast, an increase in G6PD mRNA was not observed until 12 h after refeeding starved mice and the increase was maximal (12-fold) by 27 h. Transcriptional activity of the gene was measured using the nuclear run-on assay. The G6PD probes were rigorously screened for repetitive elements and for transcription of the noncoding strand of the gene. Throughout the time course of changes in G6PD mRNA accumulation due to PUFA or refeeding, transcriptional activity of the gene did not change. Therefore, regulation of G6PD expression by nutritional status occurs at a posttranscriptional step.

Triiodothyronine-induced accumulations of malic enzyme, fatty acid synthase, acetyl-coenzyme A carboxylase, and their mRNAs are blocked by protein kinase inhibitors. Transcription is the affected step

Addition of triiodothyronine (T3) to chick-embryo hepatocytes in culture causes increased accumulations of malic enzyme, fatty acid synthase, acetyl-CoA carboxylase and their mRNAs. H-8 and other protein kinase inhibitors inhibited the T3-induced accumulations of these lipogenic enzymes and their mRNAs but had no effect on the activities of 6-phosphogluconate dehydrogenase and isocitrate dehydrogenase, enzymes not induced by T3 in chick-embryo hepatocytes. H-8 also had no effect on the activities of malic enzyme, fatty acid synthase, and acetyl-CoA carboxylase in hepatocytes not treated with T3. Synthesis of soluble protein, levels of mRNAs for beta-actin and glyceraldehyde-3-phosphate dehydrogenase, and induction of metallothionein mRNA by Zn2+ were unaffected by H-8 at concentrations that inhibited the T3-induced accumulation of lipogenic enzymes and their mRNAs. H-8 inhibited T3-induced transcription of the genes for both malic enzyme and fatty acid synthase but had little effect on transcription of the beta-actin or glyceraldehyde-3-phosphate dehydrogenase genes or on total RNA synthesis in isolated nuclei. H-8 also had no effect on binding of T3 to its nuclear receptor. In isolated nuclei, H-8 inhibited phosphorylation of total protein by 15-20%. Phosphorylation of only one major protein was consistently and substantially inhibited, indicating that the effect of H-8 was selective. These results suggest that on-going protein phosphorylation is required specifically for stimulation of transcription of the lipogenic genes by T3.

Triiodothyronine stimulates and cyclic AMP inhibits transcription of the gene for malic enzyme in chick embryo hepatocytes in culture

In chick embryo hepatocytes in culture, insulin and triiodothyronine (T3) increase malic enzyme activity and the abundance of malic enzyme mRNA by at least 50-fold, and glucagon or cAMP blocks this effect. Steps regulated by these hormones were defined by measuring transcriptional activity with the nuclear run-on assay and multiple fragments of the malic enzyme gene as probes. T3 alone caused a significant increase in transcription within 1 h, with a maximal increase of 30-40-fold occurring by 24 h. When T3 was added with insulin, 80% of the maximum rate was reached in 1 h. Insulin alone had no effect on transcription of the malic enzyme gene; it amplified the response to T3 in the first few hours after adding T3 but did not alter T3's maximal effect. Cyclic AMP for 1 h completely inhibited the increase in transcription caused by T3. The size and speed of the responses of the malic enzyme gene to T3 and cAMP suggest regulation of transcription initiation. T3-stimulated transcription of the malic enzyme gene did not require ongoing protein synthesis despite the fact that inhibitors of protein synthesis inhibited the T3-stimulated accumulation of its mRNA. T3 may directly activate transcription of this gene via its receptor. The pattern of DNase I hypersensitivity of the malic enzyme gene in chick embryo hepatocytes was the same as that in fed chick liver. Insulin, T3, and cAMP had no effect on that pattern. In chick embryo hepatocytes in culture, factors involved in regulation of transcription by insulin, T3, and cAMP may be bound to DNA independently of hormonal treatment.

Nutritional regulation and tissue-specific expression of the malic enzyme gene in the chicken. Transcriptional control and chromatin structure

Refeeding starved chicks causes a 25- to 50-fold increase in the level of malic enzyme mRNA in liver. To define the regulated steps, we measured transcriptional activity of the malic enzyme gene using the nuclear run-on assay and a variety of DNA probes specific to the malic enzyme gene. Refeeding starved chicks stimulated transcription of the malic enzyme gene in liver by 40- to 50-fold. An increased transcription rate was detectable at 1.5 h, was maximal at 3 h, and remained high at 24 h of refeeding. The level of nuclear precursor RNA for malic enzyme assessed by hybridization with intron-specific probes was high in liver of refed birds, and barely detectable in that of starved birds. These results indicate that nutritional regulation of the level of malic enzyme mRNA is transcriptional. Low levels of malic enzyme mRNA in brain, kidney, and heart correlated well with low rates of transcription of the malic enzyme gene in these tissues. In contrast to liver, neither the rate of transcription nor the steady-state level of malic enzyme mRNA was affected by refeeding starved birds. A series of DNase I-hypersensitive sites were located within 4000 base pairs upstream of the transcription start site of the malic enzyme gene in liver. The DNase I-hypersensitive region extending from the start of transcription to 400 base pairs upstream was much more pronounced in the refed state than in the starved state. This change in DNase I hypersensitivity followed the same time course as increased transcription of the malic enzyme gene. This DNase I-hypersensitive region also was present at low intensity in kidney and heart independently of nutritional state. The three constitutive DNase I-hypersensitive sites further upstream were present in liver but not in kidney or heart.

Triiodothyronine stimulates transcription of the fatty acid synthase gene in chick embryo hepatocytes in culture. Insulin and insulin-like growth factor amplify that effect

Hepatic fatty acid synthase is regulated by nutritional state. Starvation decreases and refeeding increases the activity of avian fatty acid synthase, principally by regulating transcription of the gene (Back, B. W., Goldman, M. J., Fisch, J.E., Ochs, R.A., and Goodridge, A.G. (1986) J. Biol. Chem. 261, 4190-4197). In chick embryo hepatocytes in culture, the stimulatory effect of feeding on fatty acid synthase activity is mimicked by adding triiodothyronine and insulin; the inhibitory effect of starvation is mimicked by adding glucagon or cyclic AMP. We now show that triiodothyronine alone stimulates transcription of fatty acid synthase by 4- to 6-fold, about the same as the increase in fatty acid synthase mRNA. When added alone, insulin has little or no effect on transcription, mRNA level, or enzyme activity. In combination with triiodothyronine, however, insulin amplifies the response to triiodothyronine by about 2-fold, leading to an overall increase of about 10-fold. Insulin-like growth factor 1 (IGF-1) has the same effect as insulin, no effect by itself, and amplification of the stimulation by triiodothyronine. A maximally effective dose of insulin has no effect in the presence of a maximally effective dose of IGF-1, suggesting regulation by a common pathway. It takes much less IGF-1 than insulin to achieve a given effect, suggesting that both insulin and IGF-1 may act through IGF-1 receptors. Plasma levels of IGF-1 are decreased by starvation and increased by feeding (reviewed by Froesch, E.R., and Zapf, J. (1985) Diabetologia 28, 485-493). Thus, IGF-1 may play a physiological role in the regulation of hepatic fatty acid synthase during transitions between the starved and fed states, roles previously assigned primarily to insulin and glucagon. IGF-1 regulates transcription of the fatty acid synthase gene. Insulin and IGF-1 also have similar effects on activity, mRNA abundance, and transcription of the malic enzyme gene. Glucagon or dibutyryl cyclic AMP inhibit fatty acid synthase activity and mRNA level in hepatocytes in culture by 70-80% and 60%, respectively, but have no effect on transcription of the fatty acid synthase gene, suggesting a post-transcriptional mode of regulation for cyclic AMP.

Potency of polyunsaturated and saturated fats as short-term inhibitors of hepatic lipogenesis in rats

Three dietary studies using male Sprague-Dawley rats conditioned to meal-eat a high glucose, fat-free diet and one in vitro study with isolated rat hepatocytes were designed to examine the hypothesis that polyunsaturated fats (i.e., safflower oil or linoleate) are more potent acute inhibitors of liver fatty acid synthesis than are saturated fats (i.e., beef tallow or palmitate). Fat in the first in vivo study was administered via intubation (1500 mg/rat) whereas in the second and third in vivo studies fat was added to the meal in amounts of 50, 100, 250 or 500 mg/g fat-free diet. When the rats were in a postprandial condition, significant suppression of hepatic lipogenesis required the meal to contain 38% of its energy as fat (i.e., 250 mg/g fat-free diet). At this level of fat, safflower oil was more inhibitory than beef tallow (p less than 0.05). The inhibition constant (Ki) for palmitate inhibition of fatty acid synthesis by isolated hepatocytes was fourfold greater than linoleate's Ki (fatty acid/albumin ratio of 1.4/1). When fat constituted 50% of the ingested energy, beef tallow was equivalent to safflower oil as an inhibitor of lipogenesis. Although a single meal containing 50 mg safflower oil/g fat-free diet did not decrease fatty acid synthesis, it effectively delayed the induction of lipogenesis during the first 30 min of the adaptive decrease in lipogenic enzymes attributed to polyunsaturated fats extends to short-term regulatory mechanisms.

Free fatty acid inhibition of the insulin induction of glucose-6-phosphate dehydrogenase in rat hepatocyte monolayers

Rat hepatocytes in monolayer culture were utilized to determine if the decrease in glucose-6-phosphate dehydrogenase (G6PD) activity resulting from the ingestion of fat can be mimicked by the addition of fatty acids to a chemically, hormonally defined medium. G6PD activity in cultured hepatocytes was induced several-fold by insulin. Dexamethasone or T3 did not amplify the insulin induction of G6PD. Glucose alone increased G6PD activity in cultured hepatocytes from fasted donors by nearly 500%. Insulin in combination with glucose induced G6PD an additional two-fold. The increase in G6PD activity caused by glucose was greater in hepatocytes isolated from 72 hr-fasted rats as compared to fed donor rats. Such a response was reminiscent of the "overshoot" phenomenon in which G6PD activity is induced well above the normal level by fasting-refeeding rats a high glucose diet. Addition of linoleate to the medium resulted in a significant suppression of insulin's ability to induce G6PD, but linoleate had no effect on the induction of G6PD activity by glucose alone. A shift to the right in the insulin-response curve for the induction of G6PD also was detected for the induction of malic enzyme and acetyl-CoA carboxylase. Arachidonate (0.25 mM) was a significantly more effective inhibitor of the insulin action than linoleate was. Apparently rat hepatocytes in monolayer culture can be utilized as a model to investigate the molecular mechanism by which fatty acids inhibit the production of lipogenic enzymes. In part, this mechanism of fatty acid inhibition involves desensitization of hepatocytes to the lipogenic action of insulin.

Fatty acid inhibition of hormonal induction of acetyl-coenzyme A carboxylase in hepatocyte monolayers

Primary cultures of adult rat hepatocytes were utilized to ascertain the impact of free fatty acids on the insulin plus dexamethasone induction of acetyl-CoA carboxylase. Lipogenesis was induced threefold by the combination of insulin and dexamethasone. The rise in fatty acid synthesis was accompanied by a comparable increase in the rate-determining enzyme acetyl-CoA carboxylase. Dexamethasone was required for the insulin induction of acetyl-CoA carboxylase. Under the permissive action of glucocorticoid, 10(-7) M insulin maximally increased enzyme activity. Half-maximum stimulation occurred with 5 X 10(-9) M insulin. Media containing 0.2 mM palmitate, oleate, linoleate, arachidonate, or docosahexaenoate significantly suppressed the hormonal induction of acetyl-CoA carboxylase. The extent of suppression was only 30-35% and did not vary with chain length or degree of unsaturation. Carboxylase activity was not suppressed further by raising the concentration of linoleate to 0.5 mM; however, 0.5 mM palmitate depleted the cells of ATP and abolished acetyl-CoA carboxylase activity. Therefore, based upon the inhibitory characteristics of the various fatty acids and the lack of a concentration dependency of the fatty acid inhibition, it would appear that fatty acid inhibition of the induction of acetyl-CoA carboxylase activity may not be a direct, physiological regulatory mechanism.

Role of sterol carrier protein in cholesterol metabolism

This report summarizes our recent studies on the protein known as sterol carrier protein (SCP) or fatty acid binding protein (FABP). SCP is a highly abundant, ubiquitous protein with multifunctional roles in the regulation of lipid metabolism and transport. SCP in vitro activates membrane-bound enzymes catalyzing cholesterol synthesis and metabolism, as well as those catalyzing long chain fatty acid metabolism. SCP also binds cholesterol and fatty acids with high affinity and rapidly penetrates cholesterol containing model membranes. Studies in vivo showed SCP undergoes a remarkable diurnal cycle in level and synthesis, induced by hormones and regulated in liver by translational events. SCP rapidly responds in vivo to physiological events and manipulations affecting lipid metabolism by changes in level. Thus SCP appears to be an important regulator of lipid metabolism. Preliminary evidence is presented that SCP is secreted by liver and intestine into blood and then taken up by tissues requiring SCP but incapable of adequate SCP synthesis.

Fatty acid-mediated disaggregation of acetyl-CoA carboxylase in isolated liver cells

In recent years the rapid regulation of acetyl-CoA (AcCoA) carboxylase (EC 6.4.1.2) has become of major interest because of the important role of malonyl-CoA in fatty acid synthesis, ketogenesis, and triglyceride production. AcCoA carboxylase is acutely regulated by two mechanisms: 1) phosphorylation-dephosphorylation and 2) polymer-protomer transition. Until recently polymer-protomer transition of AcCoA carboxylase in vivo has escaped detection. We developed a technique that estimates the intracellular proportion of polymer and protomer forms of AcCoA carboxylase based on the differential sensitivity of polymeric and protomeric AcCoA carboxylase to avidin inactivation. When the enzyme is in its highly aggregated conformation, the biotin prosthetic group of AcCoA carboxylase is protected from avidin binding. Thus the polymeric AcCoA carboxylase is more resistant than the protomeric conformation to avidin inactivation. Utilizing this technique with isolated liver cells we have been able to develop a model for the involvement of free fatty acids and glucagon in regulating polymer-protomer transition of AcCoA carboxylase, and the role of polymer as an intracellular determinant of AcCoA carboxylase activity. Our data suggest that the physiological regulation of AcCoA carboxylase involves the interaction of the phosphorylation mechanism with fatty acid-induced depolymerization. We propose that during periods of food deprivation the elevation in fatty acid-CoA esters promotes depolymerization of AcCoA carboxylase. In addition, glucagon induces phosphorylation of AcCoA carboxylase, which inhibits the enzyme's activity and facilitates acyl-CoA binding and depolymerization. The two separate mechanisms for regulating hepatic AcCoA carboxylase may work in concert to modulate the level of the regulatory metabolite malonyl-CoA.

Absorption and metabolism of adenine, adenosine-5'-monophosphate, adenosine and hypoxanthine by the isolated vascularly perfused rat small intestine

Intestinal vascular perfusion and in vivo live animal experiments were conducted in order to evaluate the nature and extent of the intestinal metabolism of adenine, adenosine, adenosine-5'-monophosphate (AMP) and hypoxanthine in the rat. Radiolabeled purine substrates were administered intralumenally. Intestinal contents, tissue and/or portal flow were collected and evaluated for resultant purine metabolites by liquid and paper chromatography and paper electrophoresis. Adenosine, AMP and hypoxanthine were quantitatively metabolized to end products (primarily uric acid) within 15 minutes after administration. In contrast, the metabolism of adenine to uric acid was considerably slower. Up to 20% of the administered adenine was recovered unmetabolized in the portal vasculature. Nonetheless uric acid was the primary metabolite recovered from the portal circulation in the isolated intestine regardless of the purine substrate or concentration administered. Since lumenal inosine concentrations rose sharply with increasing doses of AMP, either transport or metabolism of inosine is a rate-limiting step in the intestinal metabolism of purines to uric acid in the rat. Finally, the large percentage of the radiolabel in uric acid recovered in the lumen is consistent with the hypothesis that the intestine is an extrarenal site for purine excretion.

Graft-versus-host disease of the intestine: a protein losing enteropathy characterized by fecal alpha 1-antitrypsin

Severe hypoproteinemia often accompanies the development of graft-versus-host disease of the intestine in allogeneic bone marrow transplant patients. To determine whether or not protein loss occurs across the intestinal mucosa in this severe diarrheal illness, we measured fecal alpha 1-antitrypsin once per week in 24-h stool specimens from 25 consecutive patients during hospitalization for bone marrow transplantation. The mean alpha 1-antitrypsin concentration and serum clearance for these patients before transplantation were below 2.6 mg/g stool and 13.0 ml/day (upper limits for normals). Values for all patients increased moderately after pretransplant conditioning. Values for patients who did not develop graft-versus-host disease of the intestine returned to baseline levels; however, those for patients with graft-versus-host disease of the intestine became markedly and persistently elevated (concentration ranged from 16.6 to 51.1 mg/g, clearance from 66.6 to 384.5 ml/day). We conclude that mucosal protein exudation contributes to the hypoproteinemia of graft-versus-host disease of the intestine and that measurement of fecal alpha 1-antitrypsin can be used as a marker for this disease.