Stem cell culture engineering - process scale up and beyond

Advances in stem cell research and recent work on clinical trials employing stem cells have heightened the prospect of stem cell applications in regenerative medicine. The eventual clinical application of stem cells will require transforming cell production from laboratory practices to robust processes. Most stem cell applications will require extensive ex vivo handling of cells, from isolation, cultivation, and directed differentiation to product cell separation, cell derivation, and final formulation. Some applications require large quantities of cells in each defined batch for clinical use in multiple patients; others may be for autologous use and require only small-scale operations. All share a common requirement: the production must be robust and generate cell products of consistent quality. Unlike the established manufacturing process of recombinant protein biologics, stem cell applications will likely see greater variability in their cell source and more fluctuations in product quality. Nevertheless, in devising stem cell-based bioprocesses, much insight could be gained from the manufacturing of biological materials, including recombinant proteins and anti-viral vaccines. The key to process robustness is thus not only the control of traditional process chemical and physical variables, but also the sustenance of cells in the desired potency or differentiation state through controlling non-traditional variables, such as signaling pathway modulators.     

Extensive nucleotide variability within a 370 kb sequence from the central region of the major histocompatibility complex

The recent availability of the genomic sequence spanning the central and telomeric end of the major histocompatibility complex (MHC) has allowed a detailed study of its organisation, gene content and level of nucleotide variability. Previous analyses of nucleotide variability in the MHC have focused on the coding regions of the human leukocyte antigen (HLA) Class I and II genes. Non-coding nucleotide variability has been considered a by-product of exonic diversity. However, with the advent of genomic sequencing, the extent of non-coding nucleotide variability within the MHC has just begun to be appreciated. In this study, we compared different human haplotypes in 370 kb of sequence in the central region of the MHC to show the following: 1. unusually high levels of non-coding nucleotide variability, up to 80 times greater than elsewhere in the genome; 2. non-coding nucleotide variability greater than 1% at nucleotide sites distant to the Class I genes; 3. nucleotide variability greater than 1% maintained over regions containing highly linked loci; and 4. distinct troughs and peaks in the level of nucleotide variability. We will discuss these observations in relation to a possible role of nucleotide variability in the organisation of the MHC.     

Flexibility of the MHC class II peptide binding cleft in the bound, partially filled, and empty states: a molecular dynamics simulation study

Major histocompatibility (MHC) Class II cell surface proteins present antigenic peptides to the immune system. Class II structures in complex with peptides but not in the absence of peptide are known. Comparative molecular dynamics (MD) simulations of a Class II protein (HLA-DR3) with and without CLIP (invariant chain-associated protein) peptide were performed starting from the CLIP-bound crystal structure. Depending on the protonation of acidic residues in the P6 peptide-binding pocket the simulations stayed overall close to the start structure. The simulations without CLIP showed larger conformational fluctuations especially of alpha-helices flanking the binding cleft. Largest fluctuations without CLIP were observed in a helical segment near the peptide C-terminus binding region matching a segment recognized by antibodies specific for empty Class II proteins. Simulations on a Val86Tyr mutation that fills the peptide N-terminus binding P1 pocket or of a complex with a CLIP fragment (dipeptide) bound to P1 showed an unexpected long range effect. In both simulations the mobility not only of P1 but also of the entire binding cleft was reduced compared to simulations without CLIP. It correlates with the experimental finding that the CLIP fragment binding to P1 is sufficient to prevent antibody recognition specific for the empty form at a site distant from P1. The results suggest a mechanism how a local binding event of small peptides or of an exchange factor near P1 may promote peptide binding and exchange through a long range stabilization of the whole binding cleft in a receptive (near bound) conformation.     

Process economics evaluation of cell-free synthesis for the commercial manufacture of antibody drug conjugates

Continuous improvements of cell-free synthesis (CFS) systems have generated interest in adopting the technology for the manufacture of biologics. This paper provides an evaluation of the manufacturing cost-effectiveness of CFS for the commercial production of antibody-drug conjugates (ADCs). The evaluation was performed using an advanced techno-economic engine (TEE) built in Python. The TEE is programmed in an object-oriented environment capable of simulating a plethora of process flowsheets and predicting size and cost metrics for the process and the facility. A case study was formulated to compare the economics of whole bioprocesses based on either a CFS system or a mammalian cell system (CHO) for the manufacture of an ADC at a range of product demands. The analysis demonstrated the potential of CFS for the commercial manufacture of biologics and identified key cost drivers related to the system. The CFS system showed an approximately 80% increase in the cost of goods compared to CHO with a significant cost attributed to the in-house manufacture of the bacterial cell extract, necessary for the CFS reaction step in the process. A sensitivity and target analysis highlighted the need for further process improvements especially in the titer for the CFS process to become more competitive against well-established systems.     

Developing lisocabtagene maraleucel chimeric antigen receptor T-cell manufacturing for improved process,product quality and consistency across CD19+ hematologic indications

Background aimsAutologous chimeric antigen receptor (CAR) T-cell therapies have demonstrated substantial clinical benefit across several hematologic malignancies. However, patient-to-patient variability and heterogeneity of starting cellular material across patient populations and disease indications pose challenges to manufacturing consistency. Lisocabtagene maraleucel (liso-cel) is an autologous, CD19-directed, defined-composition, 4-1BB CAR T-cell product administered at equal target doses of CD8⁺ and CD4⁺ CAR⁺ T cells. Here the authors describe the optimization of the liso-cel manufacturing platform for product quality and consistency.MethodsLeukapheresis starting materials were collected from patients with large B-cell lymphoma, mantle cell lymphoma or chronic lymphocytic leukemia treated with liso-cel in clinical trials (NCT02631044 and NCT03331198). The liso-cel manufacturing process involves selection of CD8⁺ and CD4⁺ T cells from leukapheresis material followed by independent CD8⁺ and CD4⁺ T-cell activation, transduction, expansion, formulation and cryopreservation. Multivariate design of experimental approaches was utilized to optimize process conditions at both specific unit operations and across the process. Flow cytometry methods were used to assess cellular composition, memory phenotypes and cell proliferation. Antigen-specific functions, including cytokine secretion, cytolytic activity and proliferation, were assessed using endpoint assays after independent stimulation of CD8⁺ and CD4⁺ CAR⁺ T-cell product components.ResultsReductions in process duration time, optimization of drug product container and formulation and activation signal optimization led to significantly increased CAR⁺ T-cell product viability. The heterogeneity of patient-derived starting material, including low absolute lymphocyte counts in some samples, was reduced through early T-cell purification, leading to median T-cell frequencies >95% in selected materials across disease indications and limited non-T-cell impurities. These changes further increased lineage purity in CD8⁺ and CD4⁺ CAR⁺ T-cell drug products. CD8⁺ and CD4⁺ CAR⁺ T-cell component lot functional profiles demonstrated multifunctional mechanisms of action, including differential cytokine release, differential cytolytic kinetics and high frequencies of proliferating cells. Correlative analyses demonstrated strong underlying associations between starting material attributes and final CAR⁺ T-cell product phenotype.ConclusionsDespite substantial heterogeneity of starting leukapheresis material quality/composition between individual patients and across disease indications/histologies, the liso-cel manufacturing platform is robust and capable of generating a consistent drug product from diverse starting materials with a single manufacturing platform.     

Reshaping cell line development and CMC strategy for fast responses to pandemic outbreak

The global pandemic outbreak COVID-19 (SARS-COV-2), has prompted many pharmaceutical companies to develop vaccines and therapeutic biologics for its prevention and treatment. Most of the therapeutic biologics are common human IgG antibodies, which were identified by next-generation sequencing (NGS) with the B cells from the convalescent patients. To fight against pandemic outbreaks like COVID-19, biologics development strategies need to be optimized to speed up the timeline. Since the advent of therapeutic biologics, strategies of transfection and cell line selection have been continuously improved for greater productivity and efficiency. NGS has also been implemented for accelerated cell bank testing. These recent advances enable us to rethink and reshape the chemistry, manufacturing, and controls (CMC) strategy in order to start supplying Good Manufacturing Practices (GMP) materials for clinical trials as soon as possible. We elucidated an accelerated CMC workflow for biologics, including using GMP-compliant pool materials for phase I clinical trials, selecting the final clone with product quality similar to that of phase I materials for late-stage development and commercial production.     

Hydroxocobalamin association during cell culture results in pink therapeutic proteins

Process control of protein therapeutic manufacturing is central to ensuring the product is both safe and efficacious for patients. In this work, we investigate the cause of pink color variability in development lots of monoclonal antibody (mAb) and Fc-fusion proteins. Results show pink-colored product generated during manufacturing is due to association of hydroxocobalamin (OH-Cbl), a form of vitamin B₁₂. OH-Cbl is not part of the product manufacturing process; however we found cyanocobalamin (CN-Cbl) in cell culture media converts to OH-Cbl in the presence of light. OH-Cbl can be released from mAb and Fc-fusion proteins by conversion with potassium cyanide to CN-Cbl, which does not bind. By exploiting the differential binding of CN-Cbl and OH-Cbl, we developed a rapid and specific assay to accurately measure B₁₂ levels in purified protein. Analysis of multiple products and lots using this technique gives insight into color variability during manufacturing.     

Retroviral antibody binding of the MHC class II molecule: a biochemical influence on CD4 T cell differentiation in HIV infection?

Retroviral antibody capable of binding to the major histocompatibility complex (MHC) Class II molecule has been documented in human immunodeficiency virus-1 (HIV-1)-infected patients. Interactions between the MHC Class II receptor and the T-cell receptor (TCR) are central to the immune response. Importantly, retroviral antibody possesses a much higher binding affinity for the MHC Class II receptor, when compared to the TCR. Experiments have manipulated a number of factors related to antigen-presenting cell (APC) interaction with differentiating T-cells. These studies have observed the effects of lowering antigen dose and reducing ligand density on precursor Th (T helper) cell differentiation. Studies have also examined the effect of downregulated MHC Class II receptors and co-stimulatory molecules on APC-Th cell interaction. In addition, the sequestration of antigens away from the Class II processing pathway has been studied. These investigations reveal a general trend that can determine whether a naive CD4 T-cell becomes a Th1 or Th2-like cell. If the experimental manipulation weakens the APC-Th cell interaction, a weak ligating TCR signal results. Consequently, a weak ligating TCR signal can influence precursor Th cells to become Th2-like cells. Retroviral antibody binding of MHC Class II receptors may mimic a number of experimental conditions responsible for creating a weak ligating TCR signal.     

Accessory cell function of cells isolated from Mycobacterium leprae-induced granulomas

The large cells from Mycobacterium leprae-induced granulomas in guinea pig lymph nodes were separated by Percoll discontinuous density gradient centrifugation and on a fluorescence-activated cell sorter (FACS) using cross-reacting monoclonal antibody to human MHC Class II antigens. Large Percoll-separated cells (83% Class II antigen positive and 52% macrophage-specific antigen positive) and FACS-separated cells are able to act as antigen-presenting cells for T-cell proliferation to PPD. In previous studies, macrophage antigen-positive cells consistently failed to act as accessory cells. This indicates that there is a population of accessory cells which are macrophage antigen negative and MHC Class II antigen positive present in these M. leprae-induced granulomas.     

De-risking Pharmaceutical Tablet Manufacture Through Process Understanding,Latent Variable Modeling,and Optimization Technologies

In pharmaceutical tablet manufacturing processes, a major source of disturbance affecting drug product quality is the (lot-to-lot) variability of the incoming raw materials. A novel modeling and process optimization strategy that compensates for raw material variability is presented. The approach involves building partial least squares models that combine raw material attributes and tablet process parameters and relate these to final tablet attributes. The resulting models are used in an optimization framework to then find optimal process parameters which can satisfy all the desired requirements for the final tablet attributes, subject to the incoming raw material lots. In order to de-risk the potential (lot-to-lot) variability of raw materials on the drug product quality, the effect of raw material lot variability on the final tablet attributes was investigated using a raw material database containing a large number of lots. In this way, the raw material variability, optimal process parameter space and tablet attributes are correlated with each other and offer the opportunity of simulating a variety of changes in silico without actually performing experiments. The connectivity obtained between the three sources of variability (materials, parameters, attributes) can be considered a design space consistent with Quality by Design principles, which is defined by the ICH-Q8 guidance (USDA 2006). The effectiveness of the methodologies is illustrated through a common industrial tablet manufacturing case study.     

(Lack of) genetic diversity in immune genes predates glacial isolation in the North American mountain goat (Oreamnos americanus)

The major histocompatibility complex (MHC) plays an important role in an organism's ability to respond to pathogens. Immunogenetic diversity is advantageous as it permits the recognition of more external antigens. For this reason, MHC and immune gene variation are considered a barometer for the genetic health of wild populations. Mountain goats (Oreamnos americanus) were previously shown to have little variation at the MHC Class II Oram-DRB locus, which was attributed to population bottlenecks during the last glacial maximum (LGM). In this paper, we extended the analysis of immunogenetic variability in mountain goats to 5 genes representing the 3 classes of MHC gene (Class I OLA, Class II DRA and DRB, and Class III TNF-α) and the natural resistance-associated macrophage protein. We sequenced approximately 3000 bp from 31 individuals sampled across the range of mountain goats and found very low levels of diversity (1-3 polymorphic sites per gene) with the exception of the Class I Oram-OLA gene. Oram-OLA was nearly 30 times more diverse than the other immune genes and appears to represent a source of increased immunogenetic diversity. This diversity may be attributed to multiple loci, mediated by pathogen exposure, or potentially influenced by social factors. The distribution of SNPs was not associated with refugial history, suggesting that the current distribution of immunogenetic diversity was present prior to the LGM. These data suggest that although they have low levels of diversity at the 4 of 5 immune loci, mountain goats may be better equipped for future climate oscillations and pathogen exposure than previously thought.     

Manufacturing history of etanercept (Enbrel®): Consistency of product quality through major process revisions

Etanercept (ETN) (Enbrel®) is a soluble protein that binds to, and specifically inhibits, tumor necrosis factor (TNF), a proinflammatory cytokine. ETN is synthesized in Chinese hamster ovary cells by recombinant DNA technology as a fusion protein, with a fully human TNFRII ectodomain linked to the Fc portion of human IgG1. Successful manufacture of biologics, such as ETN, requires sophisticated process and product understanding, as well as meticulous control of operations to maintain product consistency. The objective of this evaluation was to show that the product profile of ETN drug substance (DS) has been consistent over the course of production. Multiple orthogonal biochemical analyses, which included evaluation of attributes indicative of product purity, potency, and quality, were assessed on >2,000 batches of ETN from three sites of DS manufacture, during the period 1998–2015. Based on the key quality attributes of product purity (assessed by hydrophobic interaction chromatography HPLC), binding activity (to TNF by ELISA), potency (inhibition of TNF-induced apoptosis by cell-based bioassay) and quality (N-linked oligosaccharide map), we show that the integrity of ETN DS has remained consistent over time. This consistency was maintained through three major enhancements to the initial process of manufacturing that were supported by detailed comparability assessments, and approved by the European Medicines Agency. Examination of results for all major quality attributes for ETN DS indicates a highly consistent process for over 18 years and throughout changes to the manufacturing process, without affecting safety and efficacy, as demonstrated across a wide range of clinical trials of ETN in multiple inflammatory diseases.     

Expression of major histocompatibility complex antigens on the bovine corpus luteum during the estrous cycle, luteolysis, and early pregnancy.

Treatment of cultured bovine luteal cells with the cytokine, interferon-gamma, induces the expression of Class II major histocompatibility complex antigens (MHC Ags). To determine if Class II MHC Ags are present on the CL in vivo and if the degree of Ag expression changes during luteal life span, bovine corpora lutea were obtained on Day 6, Days 10-12, and Day 18 of the estrous cycle and MHC Ag expression was evaluated via indirect immunofluorescence. Flow cytometry was used to determine the percentage of MHC Ag-positive cells on cell populations distinguished by cell size and intracellular density. Minimal Class II MHC Ag expression was detected on Day 6 CL (approximately 25%), which consisted primarily of smaller cells. The midcycle and late CL consisted of these small cells (SC) and two populations of large cells that differed in intracellular density, or right-angle light scatter. In midcycle CL, few (less than 25%) SC or large, dense cells (LDC) expressed the Class II MHC Ag whereas a high percentage (75%) of the large, less-dense cells (LLDC) were Class II MHC Ag-positive. Class II MHC Ag expression remained negligible on the LDC of the Day 18 CL; however, there was an elevation in the percentage of SC and LLDC expressing Class II Ag (p less than 0.05). To determine if Class II MHC Ag expression also varied with different functional states of the CL, bovine CL were collected after prostaglandin (PG) F2 alpha-induced regression and on Day 18 of early pregnancy. When luteolysis was allowed to progress in vivo, the percentage of Class II MHC Ag-positive cells was increased in all cell populations (p less than 0.05). Class II MHC Ag expression was significantly lower (p less than 0.05) on the three cell populations comprising the CL of pregnancy as compared to the Day 18 cyclic CL. It is hypothesized that enhanced expression of Class II MHC Ags on the late CL and during PGF2 alpha-induced regression may potentiate immune response mechanisms for luteolysis.     

A glycosylated type I membrane protein becomes cytosolic when peptide: N-glycanase is compromised

The human cytomegalovirus-encoded glycoprotein US2 catalyzes proteasomal degradation of Class I major histocompatibility complex (MHC) heavy chains (HCs) through dislocation of the latter from the endoplasmic reticulum (ER) to the cytosol. During this process, the Class I MHC HCs are deglycosylated by an N-glycanase-type activity. siRNA molecules designed to inhibit the expression of the light chain, beta(2)-microglobulin, block the dislocation of Class I MHC molecules, which implies that US2-dependent dislocation utilizes correctly folded Class I MHC molecules as a substrate. Here we demonstrate it is peptide: N-glycanase (PNGase or PNG1) that deglycosylates dislocated Class I MHC HCs. Reduction of PNGase activity by siRNA expression in US2-expressing cells inhibits deglycosylation of Class I MHC HC molecules. In PNGase siRNA-treated cells, glycosylated HCs appear in the cytosol, providing the first evidence for the presence of an intact N-linked type I membrane glycoprotein in the cytosol. N-glycanase activity is therefore not required for dislocation of glycosylated Class I MHC molecules from the ER.     

Approaches to Quality Risk Management When Using Single-Use Systems in the Manufacture of Biologics

Biologics manufacturing technology has made great progress in the last decade. One of the most promising new technologies is the single-use system, which has improved the efficiency of biologics manufacturing processes. To ensure safety of biologics when employing such single-use systems in the manufacturing process, various issues need to be considered including possible extractables/leachables and particles arising from the components used in single-use systems. Japanese pharmaceutical manufacturers, together with single-use suppliers, members of the academia and regulatory authorities have discussed the risks of using single-use systems and established control strategies for the quality assurance of biologics. In this study, we describe approaches for quality risk management when employing single-use systems in the manufacturing of biologics. We consider the potential impact of impurities related to single-use components on drug safety and the potential impact of the single-use system on other critical quality attributes as well as the stable supply of biologics. We also suggest a risk-mitigating strategy combining multiple control methods which includes the selection of appropriate single-use components, their inspections upon receipt and before releasing for use and qualification of single-use systems. Communication between suppliers of single-use systems and the users, as well as change controls in the facilities both of suppliers and users, are also important in risk-mitigating strategies. Implementing these control strategies can mitigate the risks attributed to the use of single-use systems. This study will be useful in promoting the development of biologics as well as in ensuring their safety, quality and stable supply.KEY WORDS: biologics, manufacturing technology, quality risk management, regulatory science, single-use system     

Differential expression of guinea pig class II major histocompatibility complex antigens on vascular endothelial cells in vitro and in experimental allergic encephalomyelitis

Previous studies have shown that vascular endothelial cells do not normally express major histocompatibility complex (MHC) Class II antigens either in vivo or in vitro. In this investigation it was found that endothelial in the central nervous system (CNS) of normal guinea pigs constitutively express MHC Class II antigens recognized by the monoclonal antibodies HLA-DR, 27E7, and MSgp8. This phenotype is retained when these CNS-derived endothelial cells are propagated in tissue culture. Furthermore, examination of CNS tissue taken from animals in the acute phase of chronic relapsing experimental allergic encephalomyelitis shows that additional epitopes of the MHC Class II antigen, detected by the monoclonal antibodies CI.13.1 and 22C4, are present during the diseased state. This study not only demonstrates constitutive expression of certain MHC Class II determinants by guinea pig endothelial cells, but also shows that other Class II determinants can be differentially expressed in certain disease states.     

Characterization of major histocompatibility complex class I and class II genes from the Tasmanian devil (<Emphasis Type="Italic">Sarcophilus harrisii</Emphasis>)

The Tasmanian devil (Sarcophilus harrisii) is currently threatened by an emerging wildlife disease, devil facial tumour disease. The disease is decreasing devil numbers dramatically and may lead to the extinction of the species. At present, nothing is known about the immune genes or basic immunology of the devil. In this study, we report the construction of the first genetic library for the Tasmanian devil, a spleen cDNA library, and the isolation of full-length MHC Class I and Class II genes. We describe six unique Class II beta chain sequences from at least three loci, which belong to the marsupial Class II DA gene family. We have isolated 13 unique devil Class I sequences, representing at least seven Class I loci, two of which are most likely non-classical genes. The MHC Class I sequences from the devil have little heterogeneity, indicating recent divergence. The MHC genes described here are most likely involved in antigen presentation and are an important first step for studying MHC diversity and immune response in the devil.     

Primitive endothelial cell lines from the porcine embryonic yolk sac

Endothelial cell lines have been established from cells that were isolated from porcine yolk sacs from day 18 and day 22 embryos and propagated in vitro under various growth conditions. After expansion in vitro, the general properties of the cells proved similar for the different media used. The endothelial cells expressed cell surface receptors for acetylated low-density lipoprotein and also expressed cell surface-associated angiotensin-converting enzyme. The cells showed a characteristically high level of binding for Bandeiraea simplicifolia lectin I and Dolichos biflorus agglutinin but did not bind significant amounts of Ulex europaeus lectin I. The cells expressed low but serologically detectable levels of Class I major histocompatibility complex (MHC) antigens but failed to bind antibodies directed against Class II MHC antigens. Alpha5beta1 integrins were weakly expressed, whereas vascular cell adhesion molecule-1 (CD106) and alphavbeta3 integrins were not detected. Three-dimensional tube formation was readily observed in cultures grown on Matrigel and occurred even in uncoated plastic dishes in the absence of Matrigel. In contrast to most of the adult porcine endothelial cells, yolk sac-derived endothelial cells did not possess serologically detectable receptors for porcine growth hormone (GH), an observation consistent with the finding that GH did not increase the proliferative rate of these cells. Electron microscopic examination demonstrated the presence of Weibel-Palade bodies, tight endothelial cell junctions, and typical rough endoplasmic reticulum. Exposure of the cells to either concanavalin-A-stimulated porcine splenocyte culture supernatants or to human tumor necrosis factor alpha did not cause upregulation of VCAM-1 or Class II MHC antigens. Addition of porcine interferon-gamma led to an increase in the level of expression of Class I MHC. Yolk sac endothelial cells from day 22 embryos showed a low but detectable level of expression of Class II MHC antigens, whereas the endothelial cells from day 18 embryos showed no expression of Class II antigens after interferon-gamma stimulation. The cells maintained competence to develop vascular structures in vitro and could do so after coinjection with murine tumor cells into adult, immunocompromised mice.     

Membrane specializations and endosome maturation in dendritic cells and B cells

Interest in the cell biology of antigen presentation is centered on dendritic cells (DCs) as initiators of the immune response. The ability to examine primary antigen-presenting cells, as opposed to cell lines, has opened a new window for study of antigen processing and peptide acquisition by Class II major histocompatibility complex (MHC) products, especially where intracellular trafficking of peptide-Class-II complexes is concerned. Here, we review the dynamics of Class II MHC-positive intracellular structures in dendritic cells as well as B cells. We focus on the generation of multivesicular bodies, where Class II MHC products acquire antigenic peptide, on the endosomal transport of peptide-loaded Class II MHC to the cell surface and on the importance of Class II MHC localization in membrane microdomains.     

Inhibition of endosomal proteolytic activity by leupeptin blocks surface expression of MHC class II molecules and their conversion to SDS resistance alpha beta heterodimers in endosomes.

The biosynthesis of MHC Class II molecules starts with the assembly of the alpha and beta subunits and the invariant chain. Intracellular transport of Class II molecules was followed in pulse-chase experiments of a human Epstein-Barr virus-transformed B lymphoblastoid cell line. Entry of Class II molecules into the endocytotic pathway and their cell surface appearance were monitored using neuraminidase as a fluid endocytotic marker and as a surface probe, respectively. In the course of intracellular transport, the Class II associated invariant chain is removed by proteases located in the endosomal pathway. Here, we show that leupeptin inhibits not only invariant chain breakdown, but also surface deposition of newly synthesized Class II molecules. Class II molecules display remarkable resistance to SDS at ambient temperature when occupied by peptide. We exploit this property to show that peptide binding precedes surface expression, and takes place in the course of intracellular transport through an endosomal compartment. Leupeptin blocks the conversion of Class II molecules to an SDS resistant complex.