Search

Advances in antibody discovery technologies, especially with the availability of humanized mice and phage/yeast library approaches, enable the generation of a large diversity of antibodies against nearly any target of interest. As a result, there is an increasing demand for the production of larger numbers of purified antibodies at quantities (10s-100s of milligrams) sufficient for functional screening assays, drug-ability/develop-ability studies and immunogenicity assessments. To accommodate this need, new methods are required that bridge miniature high throughput/plate-based purification and conventional, one at a time, two-step purification at much larger scales. Thus, we developed a semi-automated, mid-scale (i.e., 1–75 mg) purification process that uses a combination of parallel affinity capture and automated sequential polishing to provide substantially improved throughput while delivering high purity. We optimized the affinity capture step to perform 24 monoclonal antibody purifications in parallel using a Protein Maker for 20–200 mL culture media. The eluant is transferred directly to an AKTA pure system equipped with an autosampler for sequential preparative size exclusion chromatography to remove aggregates and undesirable impurities, as well as exchange the antibody into a buffer suitable for most uses, including cell-based assays. This two-step purification procedure, together with plate-based protein analytical methods, can purify 24–48 monoclonal antibodies in

Poor solubility is a common challenge encountered during the development of high concentration monoclonal antibody (mAb) formulations, but there are currently no methods that can provide predictive information on high-concentration behavior of mAbs in early discovery. We explored the utility of methodologies used for determining extrapolated solubility as a way to rank-order mAbs based on their relative solubility properties. We devised two approaches to accomplish this: 1) vapor diffusion technique utilized in traditional protein crystallization practice, and 2) polyethylene glycol (PEG)-induced precipitation and quantitation by turbidity. Using a variety of in-house mAbs with known high-concentration behavior, we demonstrated that both approaches exhibited reliable predictability of the relative solubility properties of these mAbs. Optimizing the latter approach, we developed a format that is capable of screening a large panel of mAbs in multiple pH and buffer conditions. This simple, material-saving, high-throughput approach enables the selection of superior molecules and optimal formulation conditions much earlier in the antibody discovery process, prior to time-consuming and material intensive high-concentration studies.

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) poses a public health threat for which preventive and therapeutic agents are urgently needed. Neutralizing antibodies are a key class of therapeutics that may bridge widespread vaccination campaigns and offer a treatment solution in populations less responsive to vaccination. Here, we report that high-throughput microfluidic screening of antigen-specific B cells led to the identification of LY-CoV555 (also known as bamlanivimab), a potent anti-spike neutralizing antibody from a hospitalized, convalescent patient with coronavirus disease 2019 (COVID-19). Biochemical, structural, and functional characterization of LY-CoV555 revealed high-affinity binding to the receptor-binding domain, angiotensin-converting enzyme 2 binding inhibition, and potent neutralizing activity. A pharmacokinetic study of LY-CoV555 conducted in cynomolgus monkeys demonstrated a mean half-life of 13 days and a clearance of 0.22 ml hour-1 kg-1, consistent with a typical human therapeutic antibody. In a rhesus macaque challenge model, prophylactic doses as low as 2.5 mg/kg reduced viral replication in the upper and lower respiratory tract in samples collected through study day 6 after viral inoculation. This antibody has entered clinical testing and is being evaluated across a spectrum of COVID-19 indications, including prevention and treatment.

LY-CoV555, a SARS-CoV-2 spike protein–specific antibody, neutralizes SARS-CoV-2 and protects the airways of nonhuman primates against infection., A monoclonal for SARS-CoV-2 Among the most promising therapeutic options for individuals with coronavirus disease 2019 (COVID-19) are monoclonal antibodies (mAbs). In this study, Jones et al. identified, characterized, and tested one such mAb, LY-CoV555, in vitro and in vivo. They found that LY-CoV555 bound to the severe acute respiratory distress syndrome coronavirus-2 (SARS-CoV-2) spike protein and prevented its interaction with angiotensin-converting enzyme 2. Prophylactic treatment with LY-CoV555 protected the upper and lower respiratory tracts of nonhuman primates from becoming infected with SARS-CoV-2. Together, these data support the clinical use of LY-CoV555 for treating patients with COVID-19., Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) poses a public health threat for which preventive and therapeutic agents are urgently needed. Neutralizing antibodies are a key class of therapeutics that may bridge widespread vaccination campaigns and offer a treatment solution in populations less responsive to vaccination. Here, we report that high-throughput microfluidic screening of antigen-specific B cells led to the identification of LY-CoV555 (also known as bamlanivimab), a potent anti-spike neutralizing antibody from a hospitalized, convalescent patient with coronavirus disease 2019 (COVID-19). Biochemical, structural, and functional characterization of LY-CoV555 revealed high-affinity binding to the receptor-binding domain, angiotensin-converting enzyme 2 binding inhibition, and potent neutralizing activity. A pharmacokinetic study of LY-CoV555 conducted in cynomolgus monkeys demonstrated a mean half-life of 13 days and a clearance of 0.22 ml hour−1 kg−1, consistent with a typical human therapeutic antibody. In a rhesus macaque challenge model, prophylactic doses as low as 2.5 mg/kg reduced viral replication in the upper and lower respiratory tract in samples collected through study day 6 after viral inoculation. This antibody has entered clinical testing and is being evaluated across a spectrum of COVID-19 indications, including prevention and treatment.

We describe a bispecific dual-antagonist antibody against human B cell activating factor (BAFF) and interleukin 17A (IL-17). An anti-IL-17 single-chain variable fragment (scFv) derived from ixekizumab (Taltz®) was fused via a glycine-rich linker to anti-BAFF tabalumab. The IgG-scFv bound both BAFF and IL-17 simultaneously with identical stoichiometry as the parental mAbs. Stability studies of the initial IgG-scFv revealed chemical degradation and aggregation not observed in either parental antibody. The anti-IL-17 scFv showed a high melting temperature (Tm) by differential scanning calorimetry (73.1°C), but also concentration-dependent, initially reversible, protein self-association. To engineer scFv stability, three parallel approaches were taken: labile complementary-determining region (CDR) residues were replaced by stable, affinity-neutral amino acids, CDR charge distribution was balanced, and a H44-L100 interface disulfide bond was introduced. The Tm of the disulfide-stabilized scFv was largely unperturbed, yet it remained monodispersed at high protein concentration. Fluorescent dye binding titrations indicated reduced solvent exposure of hydrophobic residues and decreased proteolytic susceptibility was observed, both indicative of enhanced conformational stability. Superimposition of the H44-L100 scFv (PDB id: 6NOU) and ixekizumab antigen-binding fragment (PDB id: 6NOV) crystal structures revealed nearly identical orientation of the frameworks and CDR loops. The stabilized bispecific molecule LY3090106 (tibulizumab) potently antagonized both BAFF and IL-17 in cell-based and in vivo mouse models. In cynomolgus monkey, it suppressed B cell development and survival and remained functionally intact in circulation, with a prolonged half-life. In summary, we engineered a potent bispecific antibody targeting two key cytokines involved in human autoimmunity amenable to clinical development.

Altered coagulation and inflammation contribute to the pathogenesis of ischemic renal injury. Thrombomodulin is a necessary factor in the anticoagulant protein C pathway and has inherent anti-inflammatory properties. We studied the effect of soluble thrombomodulin (sTM) in a hypoperfusion model of ischemic kidney injury. To markedly reduce infrarenal aortic blood flow and femoral arterial pressures, we clamped the suprarenal aorta of rats, occluding them 90%, for 60 min. Reversible acute kidney injury (AKI) occurred at 24 h in rats subjected to hypoperfusion. Histologic analysis at 24 h revealed acute tubular necrosis (ATN), and intravital two-photon microscopy showed flow abnormalities in the microvasculature and defects of endothelial permeability. Pretreatment with rat sTM markedly reduced both I-R-induced renal dysfunction and tubular histologic injury scores. sTM also significantly improved microvascular erythrocyte flow rates, reduced microvascular endothelial leukocyte rolling and attachment, and minimized endothelial permeability to infused fluorescence dextrans, assessed by intravital quantitative multiphoton microscopy. Furthermore, sTM administered 2 h after reperfusion protected against ischemia-induced renal dysfunction at 24 h and improved survival. By using an sTM variant, we also determined that the protective effects of sTM were independent of its ability to generate activated protein C. These data suggest that sTM may have therapeutic potential for ischemic AKI.

Administration of activated protein C (APC) protects from renal dysfunction, but the underlying mechanism is unknown. APC exerts both antithrombotic and cytoprotective properties, the latter via modulation of protease-activated receptor-1 (PAR-1) signaling. We generated APC variants to study the relative importance of the two functions of APC in a model of LPS-induced renal microvascular dysfunction. Compared with wild-type APC, the K193E variant exhibited impaired anticoagulant activity but retained the ability to mediate PAR-1-dependent signaling. In contrast, the L8W variant retained anticoagulant activity but lost its ability to modulate PAR-1. By administering wild-type APC or these mutants in a rat model of LPS-induced injury, we found that the PAR-1 agonism, but not the anticoagulant function of APC, reversed LPS-induced systemic hypotension. In contrast, both functions of APC played a role in reversing LPS-induced decreases in renal blood flow and volume, although the effects on PAR-1-dependent signaling were more potent. Regarding potential mechanisms for these findings, APC-mediated PAR-1 agonism suppressed LPS-induced increases in the vasoactive peptide adrenomedullin and infiltration of iNOS-positive leukocytes into renal tissue. However, the anticoagulant function of APC was responsible for suppressing LPS-induced stimulation of the proinflammatory mediators ACE-1, IL-6, and IL-18, perhaps accounting for its ability to modulate renal hemodynamics. Both variants reduced active caspase-3 and abrogated LPS-induced renal dysfunction and pathology. We conclude that although PAR-1 agonism is solely responsible for APC-mediated improvement in systemic hemodynamics, both functions of APC play distinct roles in attenuating the response to injury in the kidney.

The histidine-containing protein (HPr) of bacterial phosphoenolpyruvate:sugar phosphotransferase system (PTS) serves a central role in a series of phosphotransfer reactions used for the translocation of sugars across cell membranes. These studies report the high-definition solution structures of both the unphosphorylated and histidine phosphorylated (P-His) forms of HPr from Bacillus subtilis. Consistent with previous NMR studies, local conformational adjustments occur upon phosphorylation of His 15, which positions the phosphate group to serve as a hydrogen bond acceptor for the amide protons of Ala 16 and Arg 17 and to interact favorably with the alpha-helix macrodipole. However, the positively charged side chain of the highly conserved Arg 17 does not appear to interact directly with phospho-His 15, suggesting that Arg 17 plays a role in the recognition of other PTS enzymes or in phosphotransfer reactions directly. Unlike the results reported for Escherichia coli P-His HPr (Van Nuland NA, Boelens R, Scheek RM, Robillard GT, 1995, J Mol Biol 246:180-193), our data indicate that phosphorylation of His 15 is not accompanied by adoption of unfavorable backbone conformations for active site residues in B. subtilis P-Ser HPr.

Skeletal muscle wasting occurs in a great majority of cancer patients with advanced disease and is associated with a poor prognosis and decreased survival. Myostatin functions as a negative regulator of skeletal muscle mass and has recently become a therapeutic target for reducing the loss of skeletal muscle and strength associated with clinical myopathies. We generated neutralizing antibodies to myostatin to test their potential use as therapeutic agents to attenuate the skeletal muscle wasting due to cancer. We show that our neutralizing antimyostatin antibodies significantly increase body weight, skeletal muscle mass, and strength in non–tumor-bearing mice with a concomitant increase in mean myofiber area. The administration of these neutralizing antibodies in two preclinical models of cancer-induced muscle wasting (C26 colon adenocarcinoma and PC3 prostate carcinoma) resulted in a significant attenuation of the loss of muscle mass and strength with no effect on tumor growth. We also show that the skeletal muscle mass– and strength-preserving effect of the antibodies is not affected by the coadministration of gemcitabine, a common chemotherapeutic agent, in both non–tumor-bearing mice and mice bearing C26 tumors. In addition, we show that myostatin neutralization with these antibodies results in the preservation of skeletal muscle mass following reduced caloric intake, a common comorbidity associated with advanced cancer. Our findings support the use of neutralizing antimyostatin antibodies as potential therapeutics for cancer-induced muscle wasting. Mol Cancer Ther; 14(7); 1661–70. ©2015 AACR.

Solvent exchange rates and temperature coefficients for Asn/Gln side-chain amide protons have been measured in Escherichia coli HPr. The protons of the eight side-chain amide groups (two Asn and six Gln) exhibit varying exchange rates which are slower than some of the fast exchanging backbone amide protons. Differences in exchange rates of the E and Z protons of the same side-chain amide group are obtained by measuring exchange rates at pH values > 8. An NOE between a side-chain amide proton and a bound water molecule was also observed.

Catabolite repression of a number of catabolic operons in bacilli is mediated by the catabolite control protein CcpA, the phosphocarrier protein HPr from the phosphoenolpyruvate-dependent sugar transport system (PTS), and a cis-acting DNA sequence termed the catabolite-responsive element (cre). We present evidence that CcpA interacts with HPr that is phosphorylated at Ser46 (Ser(P) HPr) and that these proteins form a specific ternary complex with cre DNA. Titration experiments following the circular dichroism signal of the cre DNA indicate that this complex consists of two molecules of Ser(P) HPr, a CcpA dimer, and the cre sequence. Limited proteolysis experiments indicate that the domain structure of CcpA is similar to other members of the LacI/GalR family of helix-turn-helix proteins, comprised of a helix-turn-helix DNA domain and a C-terminal effector domain. NMR titration of Ser(P) HPr demonstrates that the isolated C-terminal domain of CcpA forms a specific complex with Ser(P) HPr but not with unphosphorylated HPr. Based upon perturbations to the NMR spectrum, we propose that the binding site of the C-terminal domain of CcpA on Ser(P) HPr forms a contiguous surface that encompasses both Ser(P)46 and His15, the site of phosphorylation by enzyme I of the PTS. This allows CcpA to recognize the phosphorylation state of HPr, effectively linking the process of sugar import via the PTS to catabolite repression in bacilli.

Leucine zipper peptides provide simple model systems for studying both the intramolecular and intermolecular interactions that govern protein folding. The synthetic 33-residue peptide GCN4-p1, derived from the yeast transcriptional activator GCN4, forms a stable biomolecular coiled-coil structure [O'Shea, E. K., Klemm, J. D., Kim, P. S., & Alber, T. (1991) Science 254, 539-544]. The guanidine-HCl induced equilibrium unfolding of this peptide at 5 degrees C and pH 7.0 yields a standard state free energy of 10.49 +/- 0.23 kcal (mol dimer)-1 when fit to a two-state model involving the native dimer and the unfolded monomer. The unfolding and refolding kinetics of GCN4-p1 were monitored by stopped-flow circular dichroism spectroscopy as a function of both peptide concentration and final denaturant concentration. The unfolding kinetics displayed single-exponential behavior, consistent with a unimolecular reaction. The refolding kinetics, which are dependent on both peptide and guanidine concentration, are well described by a simple bimolecular association reaction. A simultaneous fit of all of the unfolding and refolding kinetic data to the model, N2[symbol: see text]2U, yields refolding and unfolding rate constants in the absence of denaturant of 4.2 x 10(5) M-1 S-1 and 3.3 x 10(-3) S-1, respectively. The equilibrium unfolding curve is accurately predicted from these rate constants, providing further support for the validity of the two-state kinetic model.

Amino acid sequences in proteins can contain residues which complicate biochemical, biophysical, or protein engineering studies but which are not essential for folding or activity. Their replacement with other naturally-occurring amino acids which are not subject to such complications but which maintain essential properties of the protein is a desirable goal. A simple strategy for testing various mutants for their suitability is described for a pair of cysteine residues in dihydrofolate reductase (DHFR) from Escherichia coli. Using a reconstructed gene which preserves the amino acid sequence and introduces a variety of unique restriction sites, the cysteines at positions 85 and 152 were replaced by site-directed and cassette mutagenesis. The enzymatic activity, stability, and folding mechanism of six double mutant DHFR proteins were examined with the purpose of identifying a suitable alternative to wild type DHFR. The Cys85-->Ala and Cys152-->Ser double mutant DHFR was found to retain the four channel folding mechanism and have activity and stability which are comparable to the wild type enzyme. The replacement of the cysteines improved the resistance of DHFR to the irreversible loss of activity at high temperature.

Time-resolved fluorescence techniques were utilized to monitor the kinetic refolding reaction of Escherichia coli dihydrofolate reductase (DHFR). Measurements of emission and anisotropy decay lifetimes of both the five intrinsic tryptophan residues and the fluorescent probe 1-anilinonaphthalene-8-sulfonate (ANS) during the folding reaction were used to characterize the compactness and development of tertiary structure in transient intermediates formed during the folding of DHFR. Experiments monitoring bound ANS show that a rapidly-formed intermediate (< 20 ms) has a rotational time of approximately 10 ns and, therefore, a compactness similar to that for the native conformation. All of the tryptophan residues in this burst phase species rotate as freely as in the unfolded state. In the set of four intermediates which then appear over the next few hundred milliseconds, the apparent rotational time measured by ANS fluorescence increases to a maximum rotational time of approximately 20 ns. An increase in the average tryptophan lifetime for these intermediates suggests these side chains become excluded from solvent and associated dynamic quenching mechanisms. As the folding reaction proceeds to a set of four native conformers the bound ANS rotational time then decreases to approach that for the native protein, 10.5 ns, and the average tryptophan rotational time increases to the same value. During these rate-limiting, final steps in folding, the static quenching effects which reflect the formation of specific tertiary contacts involving tryptophans also appear.

Human follistatin is a regulatory glycoprotein with widespread biological functions including anti-inflammatory activities, wound healing properties and muscle stimulating effects. The role of follistatin in a wide range of biological activities shows promise for potential clinical application, which has prompted considerable interest in the investigation of the protein as a potential disease modifying agent. In spite of this potential, the development of follistatin as a broad use biotherapeutic has been severely hindered by a poor understanding and characterization of its pharmacokinetic and pharmacodynamic (PK/PD) relationships. Therefore, to better define these relationships, we performed in-depth analyses of the PK/PD relationships of native follistatin-315. Our data indicates that the intrinsic PK/PD properties of native follistatin-315 are poorly suited for acting as a parentally administered biotherapeutic with broad systemic effects. Herein, we leveraged protein engineering to modify the PK characteristics of the native molecule by fusing follistatin-315 to a murine IgG1 Fc and removing the intrinsic heparan sulfate binding activity of follistatin. The engineered variant molecule had ~100- and ~1600-fold improvements in terminal half-life and exposure, respectively. In contrast to the native follistatin-315, the variant showed a robust, dose-dependent pharmacological effect when administered subcutaneously on a weekly basis in mouse models of muscle atrophy and degeneration. These studies highlight the underappreciated and critical relationship between optimizing multiple physical and chemical properties of follistatin on its overall PK/PD profile. Moreover, our findings provide the first documented strategy towards the development of a follistatin therapeutic with potential utility in patients affected with skeletal muscle diseases.

The crucial step in drug discovery is the identification of a lead compound from a vast chemical library by any number of screening techniques. NMR-based screening has the advantage of directly detecting binding of a compound to the target. The spectra resulting from these screens can also be very complex and difficult to analyze, making this an inefficient process. We present here a method, RAMPED-UP NMR, (Rapid Analysis and Multiplexing of Experimentally Discriminated Uniquely Labeled Proteins using NMR) which generates simple spectra which are easy to interpret and allows several proteins to be screened simultaneously. In this method, the proteins to be screened are uniquely labeled with one amino acid type. There are several benefits derived from this unique labeling strategy: the spectra are greatly simplified, resonances that are most likely to be affected by binding are the only ones observed, and peaks that yield little or no information upon binding are eliminated, allowing the analysis of multiple proteins easily and simultaneously. We demonstrate the ability of three different proteins to be analyzed simultaneously for binding to two different ligands. This method will have significant impact in the use of NMR spectroscopy for both the lead generation and lead optimization phases of drug discovery by its ability to increase screening throughput and the ability to examine selectivity. To the best of our knowledge, this is the first time in any format that multiple proteins can be screened in one tube.

Human activated protein C (APC) is an antithrombotic, antiinflammatory serine protease that plays a central role in vascular homeostasis, and activated recombinant protein C, drotrecogin alfa (activated), has been shown to reduce mortality in patients with severe sepsis. Similar to other serine proteases, functional APC levels are regulated by the serine protease inhibitor family of proteins including α 1 -antitrypsin and protein C inhibitor. Using APC–substrate modeling, we designed and produced a number of derivatives with the goal of altering the proteolytic specificity of APC such that the variants exhibited resistance to inactivation by protein C inhibitor and α 1 -antitrypsin yet maintained their primary anticoagulant activity. Substitutions at Leu-194 were of particular interest, because they exhibited 4- to 6-fold reductions in the rate of inactivation in human plasma and substantially increased pharmacokinetic profiles compared with wild-type APC. This was achieved with minimal impairment of the anticoagulant/antithrombotic activity of APC. These data demonstrate the ability to selectively modulate substrate specificity and subsequently affect in vivo performance and suggest therapeutic opportunities for the use of protein C derivatives in disease states with elevated serine protease inhibitor levels.

The kinetic folding mechanism for Escherichia coli dihydrofolate reductase postulates two distinct types of transient intermediates. The first forms within 5 ms and has substantial secondary structure but little stability. The second is a set of four species that appear over the course of several hundred milliseconds and have secondary structure, specific tertiary structure, and significant stability (Jennings PA, Finn BE, Jones BE, Matthews CR, 1993, Biochemistry 32:3783-3789). Pulse labeling hydrogen exchange experiments were performed to determine the specific amide hydrogens in alpha-helices and beta-strands that become protected from exchange through the formation of stable hydrogen bonds during this time period. A significant degree of protection was observed for two subsets of the amide hydrogens within the dead time of this experiment (6 ms). The side chains of one subset form a continuous nonpolar strip linking six of the eight strands in the beta-sheet. The other subset corresponds to a nonpolar cluster on the opposite face of the sheet and links three of the strands and two alpha-helices. Taken together, these data demonstrate that the complex strand topology of this eight-stranded sheet can be formed correctly within 6 ms. Measurement of the protection factors at three different folding times (13 ms, 141 ms, and 500 ms) indicates that, of the 13 amide hydrogens displaying significant protection within 6 ms, 8 exhibit an increase in their protection factors from approximately 5 to approximately 50 over this time range; the remaining five exhibit protection factors > 100 at 13 ms. Only approximately half of the population of molecules form this set of stable hydrogen bonds. Thirteen additional hydrogens in the beta-sheet become protected from exchange as the set of native conformers appear, suggesting that the stabilization of this network reflects the global cooperativity of the folding reaction.

The rate-limiting steps in the folding of dihydrofolate reductase from Escherichia coli have been shown to involve the conversion of a set of four intermediates to a corresponding set of native conformers via four parallel channels [Jennings et al. (1993) Biochemistry 32, 3783-3789]. Fluorescence and absorbance studies of the unfolding and refolding of the C85S/C152E double mutant at various final urea concentrations reveal two slow folding reactions, two fewer than observed in the wild-type protein. Refolding in the presence of substoichiometric levels of the inhibitor methotrexate shows that the two remaining slow reactions correspond to two parallel channels which lead to a pair of native conformers capable of binding the inhibitor. A combination of stopped-flow circular dichroism and cofactor binding studies confirms that the four parallel channels observed in the wild-type protein have collapsed into two channels in the mutant. Kinetic and equilibrium studies of the single cysteine mutants suggest that replacements of Cysteine-85 which perturb the hydrophobic core containing this side chain are responsible for the simplification of the kinetic mechanism. These results demonstrate that at least two of the parallel folding channels in dihydrofolate reductase arise when tertiary structure develops and are not dependent upon cis/trans isomerization at prolyl peptide bonds.

The mechanism of folding of dihydrofolate reductase from Escherichia coli was reinvestigated by studying the unfolding and refolding kinetics using absorbance and fluorescence spectroscopies. The original kinetic model proposed that folding involved a series of native, intermediate, and unfolded forms which interconverted through four independent channels linked by slow cis/trans isomerization reactions at Xaa-Pro peptide bonds [Touchette, N. A., Perry, K. M., & Matthews, C. R. (1986) Biochemistry 25, 5445]. Recently, alternative sequential models have been proposed [Frieden, C. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 4413; Kuwajima et al. (1991) Biochemistry 30, 7693] which challenge the original proposal. Stopped-flow studies of the intrinsic tryptophan fluorescence demonstrated the presence of three (and tentatively four) kinetic phases in unfolding which correlated well with four phases previously observed in refolding experiments. By monitoring the binding of the inhibitor methotrexate during folding at varying relative concentrations of inhibitor to protein, it was found that the selective loss of the slow-folding phases at substoichiometric levels could only be explained by a four-channel folding model. Double-jump experiments (native-->unfolded-->native) showed that the four refolding channels are populated within 20 s at 15 degrees C and are not likely to be due to proline isomerization. Reverse double-jump experiments (unfolded-->native-->unfolded) demonstrated that interconversions between native conformers are more rapid than originally proposed. Interestingly, the majority of the protein folds through a channel to a native conformer that is minimally populated at equilibrium. This implies that although the folding of dihydrofolate reductase is ultimately under thermodynamic control, kinetic factors contribute to the transient populations of native species during folding.