New antiprogestins with partial agonist activity: potential selective progesterone receptor modulators (SPRMs) and probes for receptor- and coregulator-induced changes in progesterone receptor induction properties.

A pharmacologically relevant property of steroid hormone-regulated gene induction is the partial agonist activity of antisteroid complexes. We now report that dexamethasone-mesylate (Dex-Mes) and dexamethasone-oxetanone (Dex-Ox), each a derivative of the glucocorticoid-selective steroid dexamethasone (Dex), are two new antiprogestins with significant amounts of agonist activity with both the A and B isoforms of progesterone receptor (PR), for different progesterone-responsive elements, and in several cell lines. These compounds continue to display activity under conditions where another partial antiprogestin (RTI-020) is inactive. These new antiprogestins were used to determine whether the partial agonist activity of PR complexes can be modified by changing concentrations of receptor or coregulator, as we have recently demonstrated for glucocorticoid receptors (GRs). Because GR and coregulator concentrations simultaneously altered the position of the physiologically relevant dose-response curve, and associated EC(50), of GR-agonist complexes, we also examined this phenomenon with PR. We find that elevated PR or transcriptional intermediary factor 2 (TIF2) concentrations increase the partial agonist activity of Dex-Mes and Dex-Ox, and the EC(50) of agonists, independently of changes in total gene transactivation. Furthermore, the corepressors SMRT (silencing mediator for retinoid and thyroid receptors) and NCoR (nuclear receptor corepressor) each suppresses gene induction but NCoR acts opposite to SMRT and, like the coactivator TIF2, reduces the EC(50) and increases the partial agonist activity of antiprogestins. These comparable responses of GR and PR suggest that variations in receptor and coregulator concentrations may be a general mechanism for altering the induction properties of other steroid receptors. Finally, the magnitude of coregulator effects on PR induction properties are often not identical for agonists and the new antagonists, suggesting subtle mechanistic differences. These properties of Dex-Mes and Dex-Ox, plus the sensitivity of their activity to cellular differences in PR and coregulator concentrations, make these steroids potential new SPRMs (selective progesterone receptor modulators) that should prove useful as probes of PR induction properties.

Opposing effects of corepressor and coactivators in determining the dose-response curve of agonists, and residual agonist activity of antagonists, for glucocorticoid receptor-regulated gene expression.

A distinguishing, but unexplained, characteristic of steroid hormone action is the dose-response curve for the regulation of gene expression. We have previously reported that the dose-response curve for glucocorticoid induction of a transfected reporter gene in CV-1 and HeLa cells is repositioned in the presence of increasing concentrations of glucocorticoid receptors (GRs). This behavior is now shown to be independent of the reporter, promoter, or enhancer, consistent with our proposal that a transacting factor(s) was being titrated by added receptors. Candidate factors have been identified by the observation that changes in glucocorticoid induction parameters in CV-1 cells could be reproduced by varying the cellular levels of coactivators [transcriptional intermediary factor 2 (TIF2), steroid receptor coactivator 1 (SRC-1), and amplified in breast cancer 1 (AIB1)], comodulator [CREB-binding protein (CBP)], or corepressor [silencing mediator for retinoid and thyroid-hormone receptors (SMRT)] without concomitant increases in GR. Significantly, the effects of TIF2 and SMRT were mutually antagonistic. Similarly, additional SMRT could reverse the action of increased levels of GRs in HeLa cells, thus indicating that the effects of cofactors on transcription may be general for GR in a variety of cells. These data further indicate that GRs are yet an additional target of the corepressor SMRT. At the same time, these cofactors were found to be capable of regulating the level of residual agonist activity displayed by antiglucocorticoids. Finally, these observations suggest that a novel role for cofactors is to participate in processes that determine the dose-response curve, and partial agonist activity, of GR-steroid complexes. This new activity of cofactors is disconnected from their ability to increase or decrease GR transactivation. An equilibrium model is proposed in which the ratio of coactivator-corepressor bound to either receptor-agonist or -antagonist complexes regulates the final transcriptional properties.

Quantity of partial agonist activity for antiglucocorticoids complexed with mutant glucocorticoid receptors is constant in two different transactivation assays but not predictable from steroid structure.

An unsolved question in steroid hormone action is why the amount of agonist activity displayed by antisteroids is not constant but varies with the assay conditions. Receptor mutations have provided insight into hormone action, presumably due to changes in the tertiary structure of the receptor that alter its interaction surfaces with the transcriptional machinery or/and co-factors. We have now employed two mechanistically different induction assays to determine whether disparate transactivation processes are similarly altered by receptor mutations. The two activation assays studied were (i) the standard induction of GREtkLUC in transiently transfected CV-1 cells and (ii) a novel modulation of endogenous receptor activity by transiently transfected receptors in HeLa cells. Five different mutations in the ligand binding and DNA binding domains of the rat glucocorticoid receptor (CS1, CS1/CD, 451/9, C656G, and R732Q) and seven steroids of varied structures (five antagonists and two agonists) were selected for use. The results in both induction assays were the same. However, no generalizations regarding steroid structure and activity emerged. Neither of two potent glucocorticoids were active with GR-CS1, or GR-CS1/CD, while RU 486 was the only antisteroid with appreciable agonist activity. With the GR-451/9 mutant, three antagonists afforded partial agonist activity. We confirmed that the C656G mutant is both "super-sensitive" and "super-selective" for transactivation. In contrast, the R732Q mutation caused significant decreases in activity with both antagonists and subsaturating concentrations of agonists. This inability to generalize about the behavior of any class of steroids with mutant receptors may reflect an induced fit for each receptor steroid complex. Nevertheless, the activity of a given steroid appeared to be constant in two different transactivation assays for a given mutant receptor. Thus, disparate transactivation processes may utilize identical receptor surfaces, even in the expression of partial agonist activity for specific antiglucocorticoids.

Induction properties of a transiently transfected glucocorticoid-responsive gene vary with glucocorticoid receptor concentration.

Transient transfections of steroid receptors have yielded much of the data used to construct the current models of steroid hormone action. These experiments invariably examine the ability of receptors to regulate transcription when occupied by saturating concentrations of steroid. We now report that other induction properties of a transiently transfected gene are not constant but vary with the concentration of transiently transfected glucocorticoid receptors. Thus, the percentage of maximal induction seen with subsaturating concentrations of glucocorticoid could be dramatically increased, and an antiglucocorticoid could be converted into a partial glucocorticoid, simply by increasing the concentration of glucocorticoid receptors. This behavior was observed in HeLa cells, containing endogenous receptors, or in CV-1 cells, containing almost no endogenous receptor, with either homologous or heterologous receptors. These increases were relatively insensitive to the concentration of reporter gene, suggesting the titration of some transcription factor(s) involved in regulating the position of the glucocorticoid dose-response curve and the agonist activity of an antiglucocorticoid. This property of transfected glucocorticoid receptors required a full-length, functionally active receptor but was retained, albeit reduced in magnitude, in the absence of binding to a glucocorticoid response element. Furthermore, this phenomenon was specific in that the A form of the human progesterone receptor had no effect under the same conditions. These variations in induction properties of antiglucocorticoids and of subsaturating concentrations of glucocorticoid, in a manner that was proportional to the amount of transfected receptor, reveal processes that are not operative with saturating concentrations of glucocorticoid. These variations also demonstrate that caution should be exercised in making mechanistic conclusions based solely on experiments conducted with saturating concentrations of glucocorticoid.

The factor binding to the glucocorticoid modulatory element of the tyrosine aminotransferase gene is a novel and ubiquitous heteromeric complex.

Glucocorticoid induction of the tyrosine aminotransferase gene deviates from that of many glucocorticoid-responsive genes by having a lower EC50 and displaying more agonist activity with a given antiglucocorticoid. A cis-acting element, located 3646 base pairs upstream of the start of tyrosine aminotransferase gene transcription, has been found to be sufficient to reproduce these variations with heterologous genes and promoters (Oshima, H., and Simons, S.S., Jr. (1992) Mol. Endocrinol. 6, 416-428). This element has been called a glucocorticoid modulatory element, or GME. Others have called this sequence a cyclic AMP-responsive element (CRE) due to the binding of the cyclic AMP response element binding protein (CREB). We now report the partial purification and characterization of two new proteins (GMEB1 and -2) of 88 and 67 kDa that bind to the GME/CRE as a heteromeric complex. This purification was followed by the formation of a previously characterized, biologically relevant band in gel shift assays. By several biochemical criteria, the GMEBs differed from many of the previously described CREB/CREM/ATF family members. Partial peptide sequencing revealed that the sequences of these two proteins have not yet been described. Size exclusion chromatography and molecular weight measurements of the gel-shifted band demonstrated that the GMEBs bound to the GME as a macromolecular complex of about 550 kDa that could be dissociated by deoxycholate. Similar experiments showed that CREB bound to the GME as heteromeric complexes of about 310 and 360 kDa. As determined from gel shift assays, GMEB1 and -2 are not restricted to rat liver cells but appear to be ubiquitous. Thus, these novel GMEBs may participate in a similar modulation of other glucocorticoid-inducible genes in a variety of cells.

Microtubules are not required for glucocorticoid receptor mediated gene induction.

Steroid-free glucocorticoid receptors are generally considered to reside in the cytoplasm of cells. After the binding of steroids, the receptors translocate into the nucleus in a manner that has been proposed to involve microtubules. However, some results with inhibitors of microtubule assembly argue to the contrary. In all of these studies, only the whole cell localization of receptors has been examined; the biological activity of these receptors has not been determined. We now report that steroid-induced gene expression is maintained in the absence of intact microtubules. This argues that microtubules are not required for either the nuclear translocation or biological activity of glucocorticoid receptors.

A new cis-acting element involved in tissue-selective glucocorticoid inducibility of tyrosine aminotransferase gene expression.

Tyrosine aminotransferase (TAT), a prototypical steroid hormone-inducible gene, has been used extensively in studies of tissue-specific control of gene transcription. Over the last several years, a total of five cis-acting elements have been implicated in the tissue-specific expression and induction of the TAT gene in rat liver. These elements are all located upstream of the start of transcription, at -11, -5.5, -3.6, -2.5, and approximately -0.1 kilobases (kb). We now have used both stable and transient transfection assays to define a new element between -2.56 and -2.3 kb that regulates the fold induction by glucocorticoids in a tissue-selective manner. Compared to simple glucocorticoid-regulated constructs, which were used as controls, the major effect of this element was repression of glucocorticoid inducibility in nonliver cells. This activity was both orientation and position independent and was seen with homologous and heterologous promoters and genes. Although this element, therefore, possessed silencer-like activity, it was unable to extinguish gene expression in nonliver cells. In fact, the observance of some glucocorticoid-induced gene expression was additional evidence that the repression derived from an element that is distinct from the glucocorticoid-responsive element at -2.5 kb. A second element was found between -2.95 and -2.56 kb that acts in a tissue nonspecific manner to reduce the absolute level of gene expression in both hepatic and nonhepatic cells. The combined effects of this tissue-nonselective element and the above-mentioned tissue-selective element were to almost completely eliminate glucocorticoid inducibility in nonhepatic cells.

Modulation of the agonist activity of antisteroids by a novel cis-acting element.

The amount of agonist activity displayed by the antiglucocorticoid dexamethasone mesylate (Dex-Mes) for the induction of tyrosine aminotransferase (TAT) in rat hepatoma cells is greater than for glutamine synthetase and varies over a period of weeks. This variation, which has been reproduced over a period of 40 h by changing the density of the cells, suggests the involvement of a trans-acting factor. The target of this proposed trans-acting factor has now been localized to the region between -3.9 to -2.9 of the rat TAT gene from experiments with cells that were stably transfected with hybrid TAT/CAT constructs. Deletion experiments with transiently transfected TAT/tk promoter/CAT constructs revealed that this entire activity could be conveyed by a 21 bp sequence of the TAT gene. Gel shift experiments support the binding of a factor(s) to this 21 bp sequence. Thus the activity of the antagonist Dex-Mes is relatively independent of steroid structure and is largely determined by the further interactions of a trans-acting factor with the cis-acting sequence. We call this novel sequence a glucocorticoid modulatory element. A model is advanced which accounts for almost all of the results concerning TAT induction by glucocorticoids. This same model may also be useful in explaining why the amount of agonist activity of most antisteroids varies, even for different genes within the same cell.

Modulation of glucocorticoid induction of stably transfected tyrosine aminotransferase gene constructs involves elements up-stream of the glucocorticoid-responsive element.

Previous studies have documented that the amount of agonist activity expressed by the antiglucocorticoid dexamethasone 21-mesylate (Dex-Mes) for tyrosine aminotransferase (TAT) induction in two rat hepatoma cell lines (Fu5-5 and HTC) is greater in Fu5-5 cells and could be varied in each cell line with changes in cell density. We have proposed that both phenomena are mediated by the binding of a trans-acting factor, the concentration or activity of which is lower in HTC cells. We have now used DNase-I hypersensitivity studies to identify a possible binding site for this factor at around -3.6 kilobases (kb) of the TAT gene. Fu5-5 and HTC cells were then stably transfected with hybrid constructs either with (3.9TATCAT) or without (2.9TATCAT) this region of the TAT gene fused up-stream of a chloramphenicol acetyltransferase (CAT) reporter gene. High levels of Dex-Mes agonist activity for the induction of CAT activity in Fu5-5 cells were seen only with the 3.9TATCAT construct, indicating that the 0.97-kb region unique to this construct controlled the high levels of Dex-Mes agonist activity. Furthermore, variations in Fu5-5 cell density caused major quantitative changes in the amount of Dex-Mes agonist activity only in cells containing the 3.9TATCAT construct, consistent with the same 0.97-kb sequences also controlling the variations in Dex-Mes agonist activity. Additional studies at high and low cell densities revealed that the modulation of Dex-Mes agonist activity for both the endogenous TAT gene and the transfected TAT/CAT gene was not due to changes in the start site of gene transcription. These studies both support our previous hypothesis that modulation of Dex-Mes agonist activity results from changes in a trans-acting factor and localize a necessary cis-acting element to sequences between -3.9 and -2.9 kb of the TAT gene. These studies, thus, define a potentially new element for glucocorticoid regulation of TAT gene transcription.

Ligand-induced stimulation of epidermal growth factor receptor mutants with altered transmembrane regions.

The epidermal growth factor (EGF) receptor is a transmembrane glycoprotein composed of a large extracellular ligand-binding region connected to the cytoplasmic kinase domain by a single transmembrane (TM) region. To explore the role of the TM region in the process of receptor activation, we have generated EGF-receptor mutants with altered TM regions by utilizing in vitro site-directed mutagenesis. The TM regions of two mutant receptors were either extended (designated i626-3) or shortened (designated d625.3) by three hydrophobic amino acid residues. In the other two mutant receptors, hydrophobic amino acids were substituted by charged residues--i.e., Val-627 was replaced by glutamic acid (designated V627E) or Leu-642 was replaced by an arginine residue (designated L642R). NIH 3T3 cells lacking endogenous EGF receptors were transfected with constructs encoding either wild-type or mutant receptors and shown to express the receptor molecules using 125I-labeled EGF binding and immunoprecipitation experiments. The mutant receptors were expressed on the cell surface as polypeptides of Mr 170,000 exhibiting typical high- and low-affinity binding sites for 125I-labeled EGF. Similar to its effect on wild-type receptors, phorbol 12-myristate 13-acetate abolished the mutant-receptor high-affinity binding sites for EGF. Moreover, EGF was able to stimulate the kinase activities of wild-type and mutant receptors both in vitro and in living cells. The mutant receptors were also able to undergo EGF-induced receptor dimerization as revealed by cross-linking experiments with a bifunctional covalent cross-linking agent. These results are compatible with an intermolecular allosteric oligomerization model for receptor activation rather than with a model based on an intramolecular mechanism for receptor activation.

Kinetic parameters of the protein tyrosine kinase activity of EGF-receptor mutants with individually altered autophosphorylation sites.

Epidermal growth factor (EGF)-receptor mutants in which individual autophosphorylation sites (Tyr1068, Tyr1148 or Tyr1173) have been replaced by phenylalanine residues were expressed in NIH-3T3 cells lacking endogenous EGF-receptors. Kinetic parameters of the kinase of wild-type and mutant receptors were compared. Both wild-type and mutant EGF-receptors had a Km(ATP) 1-3 microM for the autophosphorylation reaction, and a Km(ATP) of 3-7 microM for the phosphorylation of a peptide substrate. These are similar to the Km(ATP) values reported for EGF-receptor of A431 cells. A synthetic peptide representing the major in vitro autophosphorylation site Tyr1173 of the EGF-receptor (KGSTAENAEYLRV) was phosphorylated by wild-type receptor with a Km of 110-130 microM, and the peptide inhibited autophosphorylation with a Ki of 150 microM. Mutant EGF-receptors phosphorylated the peptide substrate with a Km of 70-100 microM. A similar decrease of Km (substrate) was obtained when the phosphorylation experiments were performed with the commonly applied substrates angiotensin II and a peptide derived from c-src. The Km of angiotensin II phosphorylation was reduced from 1100 microM for wild-type receptor to 890 microM for mutant receptor and for c-src peptide from 1010 microM to 770 microM respectively. The Vmax of the kinase was dependent on receptor concentration, but was not significantly affected by the mutation. Analogs of the Tyr1173 peptide in which the tyrosine residue was replaced by either a phenylalanine or an alanine residue also inhibited autophosphorylation with Ki of 650-750 microM. These analyses show that alterations of individual autophosphorylation sites do not have a major effect on kinase activity.(ABSTRACT TRUNCATED AT 250 WORDS)

Biological activities of EGF-receptor mutants with individually altered autophosphorylation sites.

In vitro site-directed mutagenesis was used to replace individually the three known autophosphorylation sites of the epidermal growth factor (EGF)-receptor (i.e. Tyr1173, Tyr1148 and Tyr1068) by phenylalanine, a residue which cannot serve as a phosphate acceptor site. In another mutant, Tyr1173 was substituted by a serine residue. The cDNA constructs encoding either mutant or wild-type EGF-receptors were transfected into NIH-3T3 cells devoid of endogenous EGF-receptors. The mutant receptors were expressed on the cell surface and displayed typical high- and low-affinity binding sites for [125I]EGF. Phorbol ester (PMA) modulated the binding affinity of wild-type and mutant receptors in a similar manner. Mutant EGF-receptors exhibited EGF-dependent tyrosine kinase activity leading to self-phosphorylation and phosphorylation of exogenous substrates both in vitro and in living cells. The internalization and degradation of EGF-receptors were not affected by the mutations. Cells expressing mutant EGF-receptors became mitogenically responsive to EGF, indicating that none of the vital functions of the EGF-receptor were critically impaired by the loss of individual autophosphorylation sites. Maximal mitogenic stimulation correlated with the number of wild-type or mutant receptors per cell, highly expressing cells showing higher maximal stimulation. However, the dose-response curves of cells expressing mutant receptors were slightly shifted to lower concentrations of EGF, rendering the cells mitogenically responsive to lower doses of EGF than cells expressing normal EGF-receptor at similar expression levels. Basal [3H]thymidine incorporation in the presence of 0.5% calf serum was consistently higher for cells expressing mutant receptors, while the response to stimulation with 10% calf serum was not affected.

A mutant epidermal growth factor receptor with defective protein tyrosine kinase is unable to stimulate proto-oncogene expression and DNA synthesis.

Cultured NIH-3T3 cells devoid of endogenous epidermal growth factor (EGF) receptors were transfected with cDNA expression constructs encoding either normal human EGF receptor or a receptor mutated in vitro at Lys-721, a residue that is thought to function as part of the ATP-binding site of the kinase domain. Unlike the wild-type EGF-receptor expressed in these cells, which exhibited EGF-dependent protein tyrosine kinase activity, the mutant receptor lacked protein tyrosine kinase activity and was unable to undergo autophosphorylation and to phosphorylate exogenous substrates. Despite this deficiency, the mutant receptor was normally expressed on the cell surface, and it exhibited both high- and low-affinity binding sites. The addition of EGF to cells expressing wild-type receptors caused the stimulation of various responses, including enhanced expression of proto-oncogenes c-fos and c-myc, morphological changes, and stimulation of DNA synthesis. However, in cells expressing mutant receptors, EGF was unable to stimulate these responses, suggesting that the tyrosine kinase activity is essential for EGF receptor signal transduction.

Point mutation at the ATP binding site of EGF receptor abolishes protein-tyrosine kinase activity and alters cellular routing.

Cultured NIH 3T3 cells devoid of endogenous EGF receptors were transfected with cDNA constructs encoding either the human EGF receptor or an EGF receptor mutant in which Lys721, a key residue in the ATP binding site, was replaced with an alanine residue. The mutant receptor was properly processed, and it displayed both high- and low-affinity surface binding sites. Unlike the wild-type receptor, the mutant receptor did not possess intrinsic protein-tyrosine kinase activity. The initial rate of EGF internalization was similar for wild-type and mutant EGF receptors. Surprisingly, the mutant receptors were not down regulated, but appeared to recycle in transfected cells. These data suggest that degradation of normal EGF receptors after endocytosis is due to the kinase activity endogenous to this receptor. A single amino acid substitution rendered a "down-regulated" receptor into a receptor that can recycle from cytoplasmic compartment back to the cell surface.

Restriction enzyme digestions identify discrete domains in the chromatin around the promoter of the mouse alpha 2(I) collagen gene.

We have examined the chromatin structure around the +1 transcriptional start site of the mouse alpha 2(I) collagen gene by studying the accessibility of DNA to several restriction enzymes as well as to DNase I. In NIH 3T3 cells, which express high levels of alpha 2(I) collagen mRNA, we detect a DNase I-hypersensitive site from -240 to +110 relative to the start site of transcription at +1. By digesting chromatin with restriction enzymes, which cleave naked DNA at multiple sites within the -2000 to +1000 region, a considerably more complex picture was revealed. DNA sequences upstream of around -550 and downstream of +150 are much less accessible to restriction enzymes than the region between these sites and are, therefore, probably packaged in a more compact conformation. The region from around -550 to -240 although not within the DNase I-hypersensitive domain is nevertheless accessible to restriction enzymes and, therefore, presumably in a relatively "open" conformation. In addition, beginning 5' to -100 there is a gradual decrease in restriction enzyme accessibility as one approaches +150. Of particular interest is the finding that although sites at +65 and +126 are relatively accessible, a HinfI site at +113 is resistant in chromatin. In v-mos transformed NIH 3T3 cells which express alpha 2(I) collagen at much lower levels than untransformed NIH 3T3 cells, the DNase I-hypersensitive site as well as the majority of the chromatin restriction enzyme accessibility patterns are similar to those found in untransformed NIH 3T3 cells. However, a SphI site at +58 appears less accessible in the transformed cells. We also examined the chromatin of a myeloma cell line which does not synthesize alpha 2(I) collagen at detectable levels. In the nuclei of these cells the DNA of the alpha 2(I) collagen promoter is inaccessible to DNase I and to all restriction enzymes.

Pharmacokinetics of the diastereoisomers of leucovorin after intravenous and oral administration to normal subjects.

After i.v. administration of d,l-, 1-5-formyltetrahydrofolate (d,l-CHO-THF) CHO-THF was rapidly cleared from the plasma by conversion to 5-methyltetrahydrofolate (5-CH3-THF) and urinary excretion, whereas d-CHO-THF, which was not metabolized and was slowly excreted in the urine, persisted in plasma at concentrations greatly exceeding those of l-CHO-THF and 5-CH3-THF. The plasma half-life (beta) of the unnatural (d) isomer was 451 +/- 24 (S.E.) min compared to 31.6 +/- 1.1 min for the natural (l) isomer, and 227 +/- 20 min for its active metabolite, 5-CH3-THF. The half-lives and volumes of distribution of each of the three compounds were independent of dose over a range of 25 to 100 mg, indicating that mechanisms for distribution, metabolism, and excretion are not saturable over the dose range tested. The urinary clearance of l-CHO-THF or 5-CH3-THF differed only slightly from creatinine clearance, whereas urinary clearance of d-CHO-THF was only one-fifth creatinine clearance, indicating that d-CHO-THF was extensively reabsorbed. Absorption of d,l-CHO-THF after p.o. administration was stereoselective in that absorption of the l-isomer was approximately 5 times that of the d isomer. Thus, p.o. administration resulted in a more favorable ratio of active to inactive folates in plasma. At a dose of 25 mg, absorption approached 100% for l-CHO-THF compared to 20% for d-CHO-THF. However, absorption was saturable, and lower percentages of both compounds were absorbed at doses of 50 and 100 mg.

Differences in the pharmacokinetics of the diastereoisomers of citrovorum factor in dogs.

The pharmacokinetics of d- and l-citrovorum factor (CF) are quite different with respect to the postdistributional plasma decay rates. The natural (l) isomer had a half-life (beta) of 47 +/- 4 (S.E.) min compared to 143 +/- 15 min for the unnatural (d) isomer. Renal clearance was the same for both isomers and was proportional to glomerular filtration rate. Urinary excretion appeared to be the only route of elimination of the d isomer, while l-CF was extensively metabolized. Consequently, the unnatural isomer accumulated in great excess over the natural isomer and its active metabolite, 5-methyltetrahydrofolate. The apparent volume of distribution was about 58% of body weight for both isomers, which indicates that they have equal access to tissue compartments. The data suggest that d-CF can compete with l-CF and 5-methyltetrahydrofolate for entry into cells. Under certain conditions, accumulation of d-CF may interfere with rescue from methotrexate toxicity.