Epigenetic MLH1 silencing concurs with mismatch repair deficiency in sporadic, naturally occurring colorectal cancer in rhesus macaques

BACKGROUND

Naturally occurring colorectal cancers (CRC) in rhesus macaques share many features with their human counterparts and are useful models for cancer immunotherapy; but mechanistic data are lacking regarding the comparative molecular pathogenesis of these cancers.

METHODS

We conducted state-of-the-art imaging including CT and PET, clinical assessments, and pathological review of 24 rhesus macaques with naturally occurring CRC. Additionally, we molecularly characterized these tumors utilizing immunohistochemistry (IHC), microsatellite instability assays, DNAseq, transcriptomics, and developed a DNA methylation-specific qPCR assay for MLH1, CACNA1G, CDKN2A, CRABP1, and NEUROG1, human markers for CpG island methylator phenotype (CIMP). We furthermore employed Monte-Carlo simulations to in-silico model alterations in DNA topology in transcription-factor binding site-rich promoter regions upon experimentally demonstrated DNA methylation.

RESULTS

Similar cancer histology, progression patterns, and co-morbidities could be observed in rhesus as reported for human CRC patients. IHC identified loss of MLH1 and PMS2 in all cases, with functional microsatellite instability. DNA sequencing revealed the close genetic relatedness to human CRCs, including a similar mutational signature, chromosomal instability, and functionally-relevant mutations affecting KRAS (G12D), TP53 (R175H, R273*), APC, AMER1, ALK, and ARID1A. Interestingly, MLH1 mutations were rarely identified on a somatic or germline level. Transcriptomics not only corroborated the similarities of rhesus and human CRCs, but also demonstrated the significant downregulation of MLH1 but not MSH2, MSH6, or PMS2 in rhesus CRCs. Methylation-specific qPCR suggested CIMP-positivity in 9/16 rhesus CRCs, but all 16/16 exhibited significant MLH1 promoter hypermethylation. DNA hypermethylation was modelled to affect DNA topology, particularly propeller twist and roll profiles. Modelling the DNA topology of a transcription factor binding motif (TFAP2A) in the MLH1 promoter that overlapped with a methylation-specific probe, we observed significant differences in DNA topology upon experimentally shown DNA methylation. This suggests a role of transcription factor binding interference in epigenetic silencing of MLH1 in rhesus CRCs.

CONCLUSIONS

These data indicate that epigenetic silencing suppresses MLH1 transcription, induces the loss of MLH1 protein, abrogates mismatch repair, and drives genomic instability in naturally occurring CRC in rhesus macaques. We consider this spontaneous, uninduced CRC in immunocompetent, treatment-naïve rhesus macaques to be a uniquely informative model for human CRC.

Abstract A021: Clinical and molecular characterization of naturally-occurring colorectal cancer in rhesus macaques reveals mismatch repair deficiency driven by epigenetic silencing of MLH1

Naturally occurring colorectal cancers (CRC) in rhesus macaques share many features with their human counterparts and emerge at similar life stages as in human patients, but the carcinogenesis and molecular background remain elusive, limiting the value of the model for human disease and cancer therapy. In order to better understand this naturally occurring model, we established a diagnostic and staging pipeline including imaging (18F-FDG PET, plain and contrast-enhanced CT), histopathology, and clinical assessments. We observed cancer symptoms and co-morbidities in our cohort (n=16, all Indian-origin, 11 females, mean age at arrival 20.2y) such as hypoalbuminemia, fecal occult blood, and microcytic anemia, as frequently described in human CRC patients. Pathologically, all cancers were right-sided, involving the proximal colon and/or ileocecocolic junction, and most were densely fibrotic, restricting the colonic lumen. Most cancers appeared with glandular morphology and some (18.8%) had mucinous components. Immunohistochemistry revealed loss of MLH1 and PMS2 in 100% of investigated CRCs, indicating mismatch repair deficiency which furthermore resulted in microsatellite instability (PCR & fragment analysis). Whole exome sequencing revealed the close genetic relatedness to human CRCs, particularly exemplified by mutations affecting KRAS (37.5%, e.g., p.G12D), APC (31.3%), TP53 (18.8%, e.g. p.R175H), ARID1A (56.3%), and ALK (43.8%), as similarly annotated in the human COSMIC database. Nonetheless, somatic mutations do not explain the loss of MLH1 in the entire CRC cohort. Transcriptomics on the other hand revealed the transcriptional suppression of MLH1 but not MSH2, MSH6, or PMS2 in rhesus CRC in comparison to adjacent healthy colon. Moreover, comparison of differentially expressed gene sets of rhesus CRC with a human annotated database (IPA) confirmed the disease similarities observed clinically, genetically, and histopathologically. Subsequently, we investigated DNA methylation of the promoter region of MLH1 and retranslated markers for the CpG island methylator phenotype (CIMP) as described in human CRC. While only 56.3% of CRCs were considered CIMP positive (≥3/5 markers hypermethylated), 100% of investigated CRCs exhibited MLH1 promoter hypermethylation. As a result, epigenetic silencing is suggested to suppress MLH1 transcription, cause the loss of MLH1 protein, and drive mismatch repair deficiency and genomic instability in naturally occurring CRC in rhesus macaques. We therefore consider spontaneous, uninduced CRC in rhesus macaques, their treatment-naïve nature, and their unaltered immune competence an outstanding model for human disease and in particular for human cancer immunotherapy.

Citation Format: Simon Deycmar, Brendan Johnson, Karina Ray, David Caudell Caudell, John Olson, Greg Dugan, W. Shane Sills, Declan Ryan, Christopher Whitlow, Kiran K. Solingapuram Sai, Betsy Ferguson, Benjamin Bimber, Cassandra Cullin, Brandy Dozier, Emily Romero, Francois Villinger, Armando Burgos, Jeff Chou, Bruno Gomes, Michael Hettich, Maurizio Ceppi, Jehad Charo, J. Mark Cline. Clinical and molecular characterization of naturally-occurring colorectal cancer in rhesus macaques reveals mismatch repair deficiency driven by epigenetic silencing of MLH1 [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer; 2022 Oct 1-4; Portland, OR. Philadelphia (PA): AACR; Cancer Res 2022;82(23 Suppl_1):Abstract nr A021.

Abstract 1325: Naturally occurring colorectal cancer in nonhuman primates used to study human immunotherapeutic agents confirms a link between DNA methylation and mismatch repair deficiency

Background Non-human primates (NHP) such as rhesus macaques with naturally occurring cancers are a proposed model for translational cancer immunotherapy (CIT) research and have generated relevant proof-of-mechanism evidence for 3 different CIT agents. NHP spontaneously develop cancers with progression patterns, histology, and clinical symptoms similar to humans. Gene suppression by DNA hypermethylation in the promoter region is the major characteristics of the CpG-island-methylator-phenotype (CIMP) described in human CRC patients but information in rhesus macaques is scarce. To further validate these animals as translational models for CIT, we conducted a deep molecular characterization of NHP colorectal cancers and established novel qPCR panels to assess DNA methylation of marker genes published for humans.

Methods Our cohort (n=16) consisted of Indian-origin rhesus macaques (Macaca mulatta) with naturally occurring CRC (n=16, female=11). Clinical examination, imaging (contrast-enhanced CT, FDG-PET) and biopsy to confirm cancer histology were performed. Molecular characterization was done by IHC for CRC-associated mismatch repair proteins MLH1, MSH2, MSH6, and PMS2 and by PCR/electrophoresis for microsatellite instability. Ultimately, we designed DNA methylation- and rhesus-specific qPCR probes (Methylight) targeting corresponding regions as published in human patients, including CACNA1G, CDKN2A, CRABP1, MLH1, and NEUROG1 as parts of the CIMP panel and BMP3, NDRG4, and SEPTIN9 as probed for human CRC-screening.

Results MLH1 deficiency by IHC, in conjunction with PMS2 absence, is observed in all NHP CRC cases, clearly exceeding frequencies reported in human CRCs (ranging from 2-15%). Moreover, we have documented microsatellite instable cases in some NHP CRCs, analog to human CRCs. DNA methylation of the MLH1 promoter region was significantly elevated in CRCs (100% of CRCs >2-fold, p<0.0001) compared to healthy colon. We hypothesized that this elevation would suppress MLH1 mRNA expression. This hypothesis of epigenetic suppression is corroborated by both qPCR and RNA sequencing which demonstrate significantly downregulated levels of MLH1 mRNA. DNA methylation of the other markers is less consistent but revealed CIMP positive and CIMP negative cases in our NHP CRC cohort.

Conclusions Transcriptional suppression of MLH1 by promoter hypermethylation is a major and widespread driver of genetic instability and carcinogenesis in rhesus macaque colorectal cancer. Differential DNA methylation in the promoter regions as observed in NHP CRCs can provide a screening target for liquid biopsies. This work highlights the possible translatability of naturally occurring NHP cancers for human cancer immunotherapy research and can be further explored in future tumor-bearing monkey trials.

Citation Format: Simon Deycmar, Brendan Johnson, Declan Ryan, Shane Sills, David Caudell, Greg Dugan, George Schaaf, Christopher Whitlow, Kiran Kumar Solingapuram Sai, Betsy Ferguson, Benjamin Bimber, Karina Ray, Cassandra Cullin, Brandy Dozier, Armando Burgos, Michael Hettich, Bruno Gomes, Jehad Charo, Maurizio Ceppi, Mark Cline. Naturally occurring colorectal cancer in nonhuman primates used to study human immunotherapeutic agents confirms a link between DNA methylation and mismatch repair deficiency [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1325.

Phosphoinositides containing stearic acid are required for interaction between Rho GTPases and the exocyst to control the late steps of polarised exocytosis

Cell polarity is achieved by regulators such as small G proteins, exocyst members and phosphoinositides, with the latter playing a key role when bound to the exocyst proteins Sec3p and Exo70p, and Rho GTPases. This ensures asymmetric growth via the routing of proteins and lipids to the cell surface using actin cables. Previously, using a yeast mutant for a lysophosphatidylinositol acyl transferase encoded by the PSI1 gene, we demonstrated the role of stearic acid in the acyl chain of phosphoinositides in cytoskeletal organisation and secretion. Here, we use a genetic approach to characterise the effect on late steps of the secretory pathway. The constitutive overexpression of PSI1 in mutants affecting kinases involved in the phosphoinositide pathway demonstrated the role of molecular species containing stearic acid in bypassing a lack of phosphatidylinositol-4-phosphate PI(4)P at the plasma membrane, which is essential for the function of the Cdc42p module. Decreasing the levels of stearic acid-containing phosphoinositides modifies the environment of the actors involved in the control of late steps in the secretory pathway. This leads to decreased interactions between Exo70p and Sec3p, with Cdc42p, Rho1p and Rho3p, due to disruption of the GTP/GDP ratio of at least Rho1p and Rho3p GTPases, thereby preventing activation of the exocyst.

A novel non-human primate model of Pelizaeus-Merzbacher disease

Pelizaeus-Merzbacher disease (PMD) is a severe hypomyelinating disorder of the central nervous system (CNS) linked to mutations in the proteolipid protein-1 (PLP1) gene. Although there are multiple animal models of PMD, few of them fully mimic the human disease. Here, we report three spontaneous cases of male neonatal rhesus macaques with the clinical symptoms of hypomyelinating disease, including intention tremors, progressively worsening motor dysfunction, and nystagmus. These animals demonstrated a paucity of CNS myelination accompanied by reactive astrogliosis, and a lack of PLP1 expression throughout white matter. Genetic analysis revealed that these animals were related to one another and that their parents carried a rare, hemizygous missense variant in exon 5 of the PLP1 gene. These animals therefore represent the first reported non-human primate model of PMD, providing a novel and valuable opportunity for preclinical studies that aim to promote myelination in pediatric hypomyelinating diseases.

Bardet-Biedl Syndrome in rhesus macaques: A nonhuman primate model of retinitis pigmentosa

The development of therapies for retinal disorders is hampered by a lack of appropriate animal models. Higher nonhuman primates are the only animals with retinal structure similar to humans, including the presence of a macula and fovea. However, few nonhuman primate models of genetic retinal disease are known. We identified a lineage of rhesus macaques with a frameshift mutation in exon 3 of the BBS7 gene c.160delG (p.Ala54fs) that is predicted to produce a non-functional protein. In humans, mutations in this and other BBS genes cause Bardet-Biedl syndrome, a ciliopathy and a syndromic form of retinitis pigmentosa generally occurring in conjunction with kidney dysfunction, polydactyly, obesity, and/or hypogonadism. Three full- or half-sibling monkeys homozygous for the BBS7 c.160delG variant, at ages 3.5, 4 and 6 years old, displayed a combination of severe photoreceptor degeneration and progressive kidney disease. In vivo retinal imaging revealed features of severe macular degeneration, including absence of photoreceptor layers, degeneration of the retinal pigment epithelium, and retinal vasculature atrophy. Electroretinography in the 3.5-year-old case demonstrated loss of scotopic and photopic a-waves and markedly reduced and delayed b-waves. Histological assessments in the 4- and 6-year-old cases confirmed profound loss of photoreceptors and inner retinal neurons across the posterior retina, with dramatic thinning and disorganization of all cell layers, abundant microglia, absent or displaced RPE cells, and significant gliosis in the subretinal space. Retinal structure, including presence of photoreceptors, was preserved only in the far periphery. Ultrasound imaging of the kidneys revealed deranged architecture, and renal histopathology identified distorted contours with depressed, fibrotic foci and firmly adhered renal capsules; renal failure occurred in the 6-year-old case. Magnetic resonance imaging obtained in one case revealed abnormally low total brain volume and unilateral ventricular enlargement. The one male had abnormally small testes at 4 years of age, but polydactyly and obesity were not observed. Thus, monkeys homozygous for the BBS7 c.160delG variant closely mirrored several key features of the human BBS syndrome. This finding represents the first identification of a naturally-occurring nonhuman primate model of BBS, and more broadly the first such model of retinitis pigmentosa and a ciliopathy with an associated genetic mutation. This important new preclinical model will provide the basis for better understanding of disease progression and for the testing of new therapeutic options, including gene and cell-based therapies, not only for BBS but also for multiple forms of photoreceptor degeneration.

Cerebral cysts of ependymal or ventricular origin in a juvenile rhesus macaque (Macaca mulatta) with neurologic signs

Naturally occurring neurologic disease in non-human primates may be attributable to a wide-range of causes, including infectious agents, congenital or acquired malformations, degenerative diseases, and, rarely, neoplasia. We report a case of ataxia and paresis in a juvenile rhesus macaque with ependymal-lined cerebral cysts.

High speed atomic force microscopy to investigate the interactions between toxic Aβ1-42 peptides and model membranes in real time: impact of the membrane composition

Due to an aging population, neurodegenerative diseases have become a major health issue, the most common being Alzheimer's disease. The mechanisms leading to neuronal loss still remain unclear but recent studies suggest that soluble Aβ oligomers have deleterious effects on neuronal membranes. Here, high-speed atomic force microscopy was used to assess the effect of oligomeric species of a variant of Aβ1-42 amyloid peptide on model membranes with various lipid compositions. Results showed that the peptide does not interact with membranes composed of phosphatidylcholine and sphingomyelin. Ganglioside GM1, but not cholesterol, is required for the peptide to interact with the membrane. Interestingly, when they are both present, a fast disruption of the membrane was observed. It suggests that the presence of ganglioside GM1 and cholesterol in membranes promotes the interaction of the oligomeric Aβ1-42 peptide with the membrane. This interaction leads to the membrane's destruction in a few seconds. This study highlights the power of high-speed atomic force microscopy to explore lipid-protein interactions with high spatio-temporal resolution.

Interaction of Aβ1-42 peptide or their variant with model membrane of different composition probed by infrared nanospectroscopy

Toxicity of Aβ peptides involved in Alzheimer's disease is linked to the interaction of intermediate species with membranes. Nanoscale Infrared Spectroscopy enhances the study of the morphology and the secondary structure of the peptides as fibers or oligomers interacting with membranes of different compositions, with nanometer scale resolution. Membrane models are used to investigate the role of different lipids in their interactions with Aβ peptides. This work clearly brings to light that the presence of cholesterol in membranes is favorable to the interaction with Aβ peptides in oligomers or aggregates.

A tensegrity driven DNA nanopore

Control of transport across membranes, whether natural or synthetic, is fundamental in many biotechnology applications, including sensing and drug release. Mutations of naturally existing protein channels, such as hemolysin, have been explored in the past. More recently, DNA channels with conductivities in the nanosiemens range have been designed. Regulating transport across DNA channels in response to external stimuli remains an important challenge. Previous designs relied on steric hindrance to control the inner diameter of the channel, which resulted in unstable electric signatures. In this paper we introduce a new design to control electric channel conductance of a DNA nanopore. The tensegrity driven mechanism inhibits the flux of small analytes while keeping a tightly controlled ionic transport modulated by the addition of specific DNA sequences. Current signals are clearly defined, with no sign of gating, opening new perspectives in single molecule DNA sensing.

[Alpha-1 antitrypsin deficiency: A model of alteration of protein homeostasis or proteostasis]

Chronic obstructive pulmonary disease (COPD) is currently the ninth leading cause of death in France and is predicted to become the third leading cause of worldwide morbidity and mortality by 2020. Risk factors for COPD include exposure to tobacco, dusts and chemicals, asthma and alpha-1 antitrypsin deficiency. This genetic disease, significantly under-diagnosed and under-recognized, affects 1 in 2500 live births and is an important cause of lung and, occasionally, liver disease. Alpha-1 antitrypsin deficiency is a pathology of proteostasis-mediated protein folding and trafficking pathways. To date, there are only palliative therapeutic approaches for the symptoms associated with this hereditary disorder. Therefore, a more detailed understanding is required of the folding and trafficking biology governing alpha-1 antitrypsin biogenesis and its response to drugs. Here, we review the cell biological, biochemical and biophysical properties of alpha-1 antitrypsin and its variants, and we suggest that alpha-1 antitrypsin deficiency is an example of cell autonomous and non-autonomous challenges to proteostasis. Finally, we review emerging strategies that may be used to enhance the proteostasis system and protect the lung from alpha-1 antitrypsin deficiency.

Helicobacter canis colonization in sheep: a Zoonotic link

BACKGROUND

Helicobacter canis has been associated with hepatobiliary and gastrointestinal disease in dogs, cats, and humans. Infection has not been documented in other species.

MATERIALS AND METHODS

Sheep feces subjected to microaerobic culture. Isolates were characterized by genus-specific PCR, restriction fragment length polymorphism, biochemical profiling, and 16S rRNA sequence analysis.

RESULTS

Helicobacter canis was isolated from sheep feces and confirmed by the above methods. These isolates are distinct from other sheep-origin enterohepatic Helicobacter species previously isolated.

CONCLUSIONS

This study identifies sheep as H. canis reservoirs potentially important in zoonotic or foodborne transmission.

Translational errors as an early event in prion conversion

A prion is an infectious, altered form of a cellular protein which can self-propagate and affect normal phenotype. Prion conversion has been observed for mammalian and yeast proteins but molecular mechanisms that trigger this process remain unclear. Up to now, only post-translational models have been explored. In this work, we tested the hypothesis that co-translational events may be implicated in the conformation changes of the Ure2p protein of Saccharomyces cerevisiae. This protein can adopt a prion conformation leading to an [URE3] phenotype which can be easily assessed and quantified. We analyzed the effect of two antibiotics, known to affect translation, on [URE3] conversion frequency. For cells treated with G418 we observed a parallel increase of translational errors rate and frequency of [URE3] conversion. By contrast, cycloheximide which was not found to affect translational fidelity, has no influence on the induction of [URE3] phenotype. These results raise the possibility that the mechanism of prion conversion might not only involve alternative structures of strictly identical molecules but also aberrant proteins resulting from translational errors.

The protein-only theory and the yeast Saccharomyces cerevisiae: the prions and the propagons

The yeast prions represent a very attractive and tractable model for investigating the prion world. The more extensively studied yeast prion [PSI] leads to a propagation model that links auto-aggregation in amyloid formation and inactivation of the cellular function of the yeast 'prion protein' Sup35p. The other prion model, [URE3], appears to be similar in some genetic and biochemical properties. The characterisation of both Sup35p and Ure2p, the two 'prion proteins', mainly focusing on their aggregation properties, support this model. However, some important differences still exist that should be examined carefully. In particular, we have shown that Ure2p aggregation in vivo (monitored by fluorescence of Ure2-GFP fusion) does not necessarily give rise to a [URE3] phenotype. Comparisons of these two systems as well as more recent experiments are discussed in this review.

Stability, folding, dimerization, and assembly properties of the yeast prion Ure2p

The [URE3] factor of Saccharomyces cerevisiae propagates by a prion-like mechanism and corresponds to the loss of the function of the cellular protein Ure2. The molecular basis of the propagation of this phenotype is unknown. We recently expressed Ure2p in Escherichia coli and demonstrated that the N-terminal region of the protein is flexible and unstructured, while its C-terminal region is compactly folded. Ure2p oligomerizes in solution to form mainly dimers that assemble into fibrils [Thual et al. (1999) J. Biol. Chem. 274, 13666-13674]. To determine the role played by each domain of Ure2p in the overall properties of the protein, specifically, its stability, conformation, and capacity to assemble into fibrils, we have further analyzed the properties of Ure2p N- and C-terminal regions. We show here that Ure2p dimerizes through its C-terminal region. We also show that the N-terminal region is essential for directing the assembly of the protein into a particular pathway that yields amyloid fibrils. A full-length Ure2p variant that possesses an additional tryptophan residue in its N-terminal moiety was generated to follow conformational changes affecting this domain. Comparison of the overall conformation, folding, and unfolding properties, and the behavior upon proteolytic treatments of full-length Ure2p, Ure2pW37 variant, and Ure2p C-terminal fragment reveals that Ure2p N-terminal domain confers no additional stability to the protein. This study reveals the existence of a stable unfolding intermediate of Ure2p under conditions where the protein assembles into amyloid fibrils. Our results contradict the intramolecular interaction between the N- and C-terminal moieties of Ure2p and the single unfolding transitions reported in a number of previous studies.

The yeast prion [URE3] can be greatly induced by a functional mutated URE2 allele

The non-Mendelian element [URE3] of yeast is considered to be a prion form of the Ure2 protein. The [URE3] phenotype occurs at a frequency of 10(-5) in haploid yeast strains, is reversible, and its frequency is increased by overexpressing the URE2 gene. We created a new mutant of the Ure2 protein, called H2p, which results in a 1000-fold increase in the rate of [URE3] occurrence. To date, only the overexpression of various C-terminal truncated mutants of Ure2p gives rise to a comparable level. The h2 allele is, thus, the first characterized URE2 allele that induces prion formation when expressed at a low level. By shuffling mutated and wild-type domains of URE2, we also created the first mutant Ure2 protein that is functional and induces prion formation. We demonstrate that the domains of URE2 function synergistically in cis to induce [URE3] formation, which highlights the importance of intramolecular interactions in Ure2p folding. Additionally, we show using a green fluorescent protein (GFP) fusion protein that the h2 allele exhibits numerous filiform structures that are not generated by the wild-type protein.

The [URE3] yeast prion: from genetics to biochemistry

[URE3] is a non-Mendelian genetic element of the yeast Saccharomyces cerevisiae, an altered prion form of Ure2 protein. We show that recombinant Ure2p is a soluble protein that can assemble in vitro into dimers, tetramers, and octamers or form insoluble fibrils observed for PrP in its filamentous form or for Sup35p upon self-assembling, suggesting a similar mechanism for all prions. Computational, genetic, biochemical, and structural data allow us to specify a new boundary between the so-called prion-forming and nitrogen regulator (catalytic) domains of the protein and to map this boundary to Met-94. We bring strong evidence that the COOH-terminal (94-354) part of the protein forms a tightly folded domain, while the NH2-terminal (1-94) part is unstructured. These domains (or various parts of these domains) were shown (by means of the two-hybrid system approach and affinity binding experiments) to interact with each other (both in vivo and in vitro). We bring also evidence that the COOH-terminal (94-354) catalytically active part of the protein can be synthesized (both in vitro and in vivo) via an internal ribosome-binding mechanism, independently of the production of the full-length protein. We finally show that Ure2p aggregation in vivo (monitored by fluorescence of Ure2p--GFP fusion) does not necessarily give rise to [URE3] phenotype. The significance of these findings for the appearance and propagation of the yeast prion [URE3] is discussed.

Structural characterization of Saccharomyces cerevisiae prion-like protein Ure2

Sacchromyces cerevisiae prion-like protein Ure2 was expressed in Escherichia coli and was purified to homogeneity. We show here that Ure2p is a soluble protein that can assemble into fibers that are similar to the fibers observed in the case of PrP in its scrapie prion filaments form or that form on Sup35 self-assembly. Ure2p self-assembly is a cooperative process where one can distinguish a lag phase followed by an elongation phase preceding a plateau. A combination of size exclusion chromatography, sedimentation velocity, and electron microscopy demonstrates that the soluble form of Ure2p consists at least of three forms of the protein as follows: a monomeric, dimeric, and tetrameric form whose abundance is concentration-dependent. By the use of limited proteolysis, intrinsic fluorescence, and circular dichroism measurements, we bring strong evidence for the existence of at least two structural domains in Ure2p molecules. Indeed, Ure2p NH2-terminal region is found poorly structured, whereas its COOH-terminal domain appears to be compactly folded. Finally, we show that only slight conformational changes accompany Ure2p assembly into insoluble high molecular weight oligomers. These changes essentially affect the COOH-terminal part of the molecule. The properties of Ure2p are compared in the discussion to that of other prion-like proteins such as Sup35 and mammalian prion protein PrP.

Characterization of the interaction domains of Ure2p, a prion-like protein of yeast

In the yeast Saccharomyces cerevisiae, the non-Mendelian inherited genetic element [URE3] behaves as a prion. A hypothesis has been put forward which states that [URE3] arises spontaneously from its cellular isoform Ure2p (the product of the URE2 gene), and propagates through interactions of the N-terminal domain of the protein, thus leading to its aggregation and loss of function. In the present study, various N- and C-terminal deletion mutants of Ure2p were constructed and their cross-interactions were tested in vitro and in vivo using affinity binding and a two-hybrid analysis. We show that the self-interaction of the protein is mediated by at least two domains, corresponding to the first third of the protein (the so-called prion-forming domain) and the C-terminal catalytic domain.

Enhanced expression of the yeast Ure2 protein in Escherichia coli: the effect of synonymous codon substitutions at a selected place in the gene

The expression of the yeast Ure2 protein and its two N- and C-terminal HA-(YPYPVDYA) epitope and His-tag fusions has been enhanced in E. coli by selected silent mutagenesis of the URE2 gene. The two Arg-AGA codons at positions 253 and 254 of the URE2 gene coding sequence were exchanged by CGT codons accordingly. This has allowed an increased yield (up to 100-fold) of the full-length protein synthesized. Western blotting with HA-epitope-specific antibodies using N- and C-terminal Ure2p-HA(epitope)-His-tag fusion constructs confirmed the integrity of the recombinant proteins. The N-(C-) terminal tagged proteins were shown to possess biological activity of the natural Ure2 protein.

Differential resistance to proteinase K digestion of the yeast prion-like (Ure2p) protein synthesized in vitro in wheat germ extract and rabbit reticulocyte lysate cell-free translation systems

The Ure2p yeast prion-like protein was translated in vitro in the presence of labeled [35S]methionine in either rabbit reticulocyte lysate (RRL) or wheat germ extract (WGE) cell-free systems. When subjected to proteinase K digestion, the Ure2p protein synthesized in WGE was proteolysed much more slowly compared to that synthesized in RRL; this displays fragments of about 31-34 kDa, persisting over 8 min. Thus, the digestion rate and pattern of the protein synthesized in WGE, unlike that synthesized in RRL, revealed characteristic features of the [URE3] prion-like isoform of the Ure2p protein [Masison, D.C. and Wickner, R.B. (1995) Science 270, 93-95]. Chloramphenicol acetyltransferase, synthesized under the same conditions, differed fundamentally in its proteolytic sensitivity toward proteinase K (PK); in the RRL system it was more slowly digested than in WGE, proving specific PK inhibitors to be absent in both systems. Posttranslational addition of the WGE to the RRL-synthesized Ure2p does not protect Ure2p from efficient PK degradation either. The differences in Ure2p degradation may be ascribed to a specific structure or specific states of association of Ure2p synthesized in WGE; obviously, they yield a protein that mimics the behavior of the Ure2p in [URE3] yeast strains. The present data suggest that particular conditions of the Ure2p protein translation and/or certain cellular components (accessory proteins and extrinsic factors), as well as the nature of the translation process itself, could affect the intracellular folding pathway of Ure2p leading to the de novo formation of the prion [URE3] isoform.

Functional analysis of Rrp7p, an essential yeast protein involved in pre-rRNA processing and ribosome assembly

During the functional analysis of open reading frames (ORFs) identified during the sequencing of chromosome III of Saccharomyces cerevisiae, the previously uncharacterized ORF YCL031C (now designated RRP7) was deleted. RRP7 is essential for cell viability, and a conditional null allele was therefore constructed, by placing its expression under the control of a regulated GAL promoter. Genetic depletion of Rrp7p inhibited the pre-rRNA processing steps that lead to the production of the 20S pre-rRNA, resulting in reduced synthesis of the 18S rRNA and a reduced ratio of 40S to 60S ribosomal subunits. A screen for multicopy suppressors of the lethality of the GAL::rrp7 allele isolated the two genes encoding a previously unidentified ribosomal protein (r-protein) that is highly homologous to the rat r-protein S27. When present in multiple copies, either gene can suppress the lethality of an RRP7 deletion mutation and can partially restore the ribosomal subunit ratio in Rrp7p-depleted cells. Deletion of both r-protein genes is lethal; deletion of either single gene has an effect on pre-rRNA processing similar to that of Rrp7p depletion. We believe that Rrp7p is required for correct assembly of rpS27 into the preribosomal particle, with the inhibition of pre-rRNA processing appearing as a consequence of this defect.

Construction of a yeast strain deleted for the TRP1 promoter and coding region that enhances the efficiency of the polymerase chain reaction-disruption method

The sequence of the genome of Saccharomyces cerevisiae was recently determined. As well as all the informations concerning the structure of the chromosomes the scientific community had to deal with the discovery of dozens of new open reading frames (ORFs) of unknown function. The study of these ORFs requires the development of simple procedures that can be used on a large scale. In the framework of a European Pilot Project we have described a new approach for deleting ORFs. This method is based on transformation with a polymerase chain reaction product but is limited by the use of a strain deleted for the auxotropic marker. We present here the construction of a new recipient strain that lacks the TRP1 region and that allows a high efficiency of gene deletion.

Functional analysis of YCL09C: evidence for a role as the regulatory subunit of acetolactate synthase

We have analysed the function of the open reading frame (ORF) YCL09C. The deletion of this ORF from chromosome III does not affect the physiology of the corresponding yeast strain enough to give a distinct phenotype. Nevertheless a computational analysis reveals high homology between this ORF and the enterobacterial genes encoding the regulatory subunit of acetolactate synthase. We have therefore tested the possibility that yc109cp is the regulatory subunit of yeast acetolactate synthase by in vitro enzymatic analysis. The acetolactate synthase was previously shown to be retroinhibited by its final product valine. In Escherichia coli this retro-control is assured by the regulatory subunit. Using a yeast strain carrying a complete deletion of YCL09C, we have observed the loss of such retro-inhibition. These results together with the computational predictions show that YCL09C encodes the regulatory subunit of yeast acetolactate synthase.

Cloning and characterization of a yeast cytochrome b5-encoding gene which suppresses ketoconazole hypersensitivity in a NADPH-P-450 reductase-deficient strain

Cytochrome P-450 (Cyp) 51 or lanosterol-C14-demethylase is the main target for antifungal compounds of the triazole family like ketoconazole (Kz). Disruption of the associated NADPH-P-450 reductase-encoding gene (YRED) is not lethal, but decreases by about 20-fold the Kz resistance (KzR) of wild-type (wt) Saccharomyces cerevisiae. Transformation of a YRED-disrupted strain by a yeast genomic library based on a multicopy vector allowed us to identify a suppressor of Kz hypersensitivity. Deletion analysis of the 5-kb cloned fragment indicated that yeast cytochrome b5-encoding gene (CYB5), which encodes a 120-amino-acid (aa) protein, is required and sufficient for the suppressor effect. The encoded polypeptide shares about 30% aa identity with mammalian cytochromes b5 (Cyb5). CYB5 disruption and tetrad analysis demonstrate that yeast Cyb5 is not required for growth in a Yred+ strain. Determination of the microsomal content of b-type cytochromes by differential spectra indicated the presence of a strongly decreased or null Cyb5 level in the disrupted strain. This confirms that we have cloned the gene encoding the major microsomal form of Cyb5 which appears not to be essential. Minor Cyb5 isoforms could also be present in yeast or other redox proteins could substitute for the pleiotropic roles of Cyb5 in the sterol and lipid biosynthesis pathways.

Cellular localization of RNA14p and RNA15p, two yeast proteins involved in mRNA stability

RNA14 and RNA15 were originally identified by temperature-sensitive mutations that cause a rapid decrease in poly(A)-tail length and overall mRNA levels at the restrictive temperature. We have raised antibodies to the RNA14 and RNA15 proteins, and used subcellular fractionation and immunofluorescence to localize these proteins within the yeast cell. RNA14p is a 73 kDa protein found in both the nucleus and the cytoplasm, whilst RNA15p is a 42 kDa protein detected only in the nucleus. The observation that both proteins are found in the nucleus is in agreement with previous genetic data which suggest an interaction between RNA14p and RNA15p. Also the joint nuclear localization is consistent with the biochemical data suggesting a role in polyadenylation. The detection of significant amounts of RNA14p in the cytoplasm opens the possibility of a second function for this protein, either in cytoplasmic regulation of mRNA deadenylation or, more interestingly, in mRNA stability.

The sequence of 29.7 kb from the right arm of chromosome II reveals 13 complete open reading frames, of which ten correspond to new genes

We have determined the complete nucleotide sequence of a 29.7 kb segment from the right arm of chromosome II carried by the cosmid alpha 61. The sequence encodes the 3' region of the IRA1 gene and 13 complete open reading frames, of which ten correspond to new genes and three (CIF1, ATPsv and CKS1) have been sequenced previously. The density of protein coding sequences is particularly high and corresponds to 84% of the total length. Two new genes encode membrane proteins, one of which is particularly large, 273 kDa. In one case (ATPsv), the comparison of our sequence and the published sequence reveals significant differences.

Multipurpose vectors designed for the fast generation of N- or C-terminal epitope-tagged proteins

In this paper are described a set of new high-copy-number yeast vectors, which are specially designed for the conditional expression of epitope-tagged proteins in vivo. One of the major advantages of these plasmids is that they allow polymerase chain reaction-amplified open reading frames to be automatically fused in frame with the epitope-coding sequence, avoiding longer procedures such as site-directed mutagenesis. This heterologous construction can be realized either at the 5'-end of the coding sequence, in the pYeF1 vector, or at its 3'-end, in pYeF2, generating N- or C-terminal tagged proteins, respectively. Moreover, to increase the usefulness of the method, derivatives of the two basic URA3-borne pYeF1 and pYeF2 were constructed, carrying either the HIS3 or TRP1 gene as a marker of selection. These vectors could be of use for the purpose of functional analysis of the newly discovered genes resulting from the systematic sequencing of the yeast genome. Here, we present results showing the functional expression and the efficient immunoprecipitation of the epitope-tagged Rna15 protein, which is involved in Saccharomyces cerevisiae mRNA stability.

Expression in yeast of three allelic cDNAs coding for human liver P-450 3A4. Different stabilities, binding properties and catalytic activities of the yeast-produced enzymes

Three natural allelic cDNAs coding for P-450 3A4, the major form in human liver, namely NF25, NF10 and hPCN1, have been expressed in Saccharomyces cerevisiae. NF25 and hPCN1 were functionally expressed in yeast microsomes, yielding proteins with an absorption maximum at 448 nm in the CO-reduced difference spectrum. Some catalytic activities and substrate binding properties of P-450 NF25 and P-450 hPCN1 in yeast microsomes have been compared; no striking difference was found, showing that the two point substitutions between their amino-acid sequences (Trp392 and Thr431 in P-450 NF25 are replaced by Val392 and Ile431 in P-450 hPCN1) have no significant effect on the functional properties of these two variants. By contrast, P-450 NF10, which differs from P-450 NF25 by a one-amino-acid deletion (Ile224 replacing Thr224-Val225), was produced as a denatured form, as revealed by an absorption maximum at 420 nm, and was not catalytically active. This suggests that the deletion prevents the correct folding of the protein. The results of this study show that P-450 NF25 and P-450 hPCN1 are two roughly equivalent, functionally active variants of P-450 3A4, but that P-450 NF10 is a defective, unstable gene product that could arise from an alternative mRNA splicing. This could contribute to the large variations reported for nifedipine oxidation, a typical P-450 3A4 activity, in human liver.

Recombinant yeast in drug metabolism

The usefulness of cDNA-directed expression of human hepatic P450s in yeast for the in vitro study of drug metabolism is emphasized. The major advantages of yeast expression are: (i) relatively high yields of heterologous P450 (approximately 5-10 nmol/l of culture medium) can be obtained; (ii) the expressed P450s are directly active in yeast microsomes, allowing the determination of specific catalytic activities of individual isoforms, which is a prerequisite for the prediction of metabolic pathways for new drug candidates; (iii) transformed yeast microsomes can also be used to study the specific affinity of individual P450s for various substrates and the formation of P450-metabolite complexes by difference visible spectroscopy; such studies can help to predict drug interactions. The advantages of expression in yeast with respect to biochemical studies of drug metabolism are illustrated with data about P450 NF25 (P450 3A4), the major form of human liver. Expressed P450 NF25 is obtained in a functionally active state, and some specific catalytic activities observed in liver microsomes could be reproduced directly with transformed yeast microsomes. The use of genomically modified yeast strains coexpressing human cytochrome b5 and/or overexpressing yeast P450-reductase allowed us to optimize these catalytic activities. In particular, this coexpression system was useful in the study of the in vitro formation of a P450 NF25 Fe(II)-RNO complex. Such inhibitory complexes have been implied in numerous drug interactions involving P450 3A4.

Enhanced in vivo monooxygenase activities of mammalian P450s in engineered yeast cells producing high levels of NADPH-P450 reductase and human cytochrome b5

We have engineered yeast genomic DNA to construct a set of strains producing various relative amounts of yeast NADPH-P450 reductase (Yred) and human cytochrome b5 (Hb5). Expression of cDNAs encoding human P450 1A1, 1A2, 3A4, 19A and mouse P450 1A1 in the different oxido-reduction backgrounds thus constituted were achieved after strain transformation by plasmid-based P450-encoding expression cassettes. The results indicate that the level of Yred strongly affects all activities tested. In contrast, the amount of Hb5 affects activities in a manner that is dependent both on the P450 isoform considered and the Yred level. In a strain containing optimized amounts of Hb5 and Yred, human P450 3A4-specific testosterone-6 beta-hydroxylase activity can be enhanced as much as 73-fold in comparison with the activity observed in a wild-type strain. Bioconversion of sterols or xenobiotics was easily achieved in vivo using this new co-expression system.

Two distinct sequences control the targeting and anchoring of the mouse P450 1A1 into the yeast endoplasmic reticulum membrane

We previously expressed mouse P450 1A1 in the yeast S. cerevisiae. In the present study, I describe experiments in which several deletions in the 5' end of the corresponding cDNA were created. The truncated forms were then expressed in yeast cells. Studies of microsomes obtained from transformed yeast show that the signal-sequence is not required in vivo for the integration of mouse P450 1A1 into the endoplasmic reticulum membrane. In addition, the cytochrome deleted for its hydrophobic signal-sequence appears to be enzymatically functional. These results strongly argue for the existence of a second determinant of membrane targeting and binding.

Expression of human liver cytochrome P450 IIIA4 in yeast. A functional model for the hepatic enzyme

Cytochrome P-450 (P450) NF, a member of the P450 IIIA subfamily, is the major contributor to the oxidation of the calcium-channel blocker nifedipine in human liver microsomes. A cDNA clone designated NF25 encoding for human P450 NF was isolated from a bacteriophage lambda gt11 expression library [Beaune, P. H., Umbenhauer, D. R., Bork, R. W., Lloyd, R. S. & Guengerich, F. P. (1986) Proc. Natl Acad. Sci. USA 83, 8064-8068]. We have expressed NF25 cDNA in Saccharomyces cerevisiae using an expression vector constructed from pYeDP1/8-2 [Cullin, C. & Pompon, D. (1988) Gene 65, 203-217]. Yeast transformed with the plasmid containing the NF25 sequence (pVNF25) showed a ferrous-CO spectrum typical of cytochrome P-450. Microsomal preparations contained a protein with an apparent molecular mass identical to that of P450-5 (a form isolated from human liver indistinguishable from P450 NF) that was not present in microsomes from control yeast (transformed with pYeDP1/8-2 alone), as revealed by immunoblotting with anti-P450-5 antibodies. On the other hand, antibodies raised in rabbits against human liver P450 IIC8-10 and rat liver P450 IA1 and P450 IIE1 did not recognize yeast-expressed P450 NF25. The P450 NF25 content in microsomes was about 90 pmol/mg protein. Microsomal, yeast-expressed P450 NF25 exhibited a high affinity for different substrates including macrolide antibiotics, dihydroergotamine and miconazole as shown by difference visible spectroscopy. Microsomal suspensions containing P450 NF25 were also able to catalyze several oxidation reactions that were expected from the activities of the protein isolated from human liver, including nifedipine 1,4-oxidation, quinidine 3-hydroxylation and N-oxygenation, and N-demethylation of the macrolide antibiotics erythromycin and troleandomycin. The yeast endogenous NADPH-cytochrome P-450 reductase thus couples efficiently with the heterologous P450 NF25 though its level is far lower than that of its ortholog in human liver. Indeed addition of rabbit liver NADPH-cytochrome P-450 reductase increased the oxidation rates. Rabbit liver cytochrome b5 also caused a marked enhancement of catalytic activities, as had been noted previously for this particular P450 enzyme in a reconstituted system involving the protein purified from human liver. Furthermore, the level of the yeast endogenous cytochrome P-450 (lanosterol 14-demethylase) has been found to be negligible compared to the heterologously expressed cytochrome P-450 (30 times less). Thus, yeast microsomes containing P450 NF25 constitute by themselves a good functional model for studying the binding capacities and catalytic activities of this individual form of human hepatic cytochrome P-450.

Maximizing the expression of mammalian cytochrome P-450 monooxygenase activities in yeast cells

Cytochrome P-450s constitute a superfamily of mono-oxygenases which require the association with specific redox enzymes bound to the endoplasmic reticulum membrane for their activity. Conditions for the functional expression of these mammalian enzymes in yeast cells and the respective merits and limitations of currently used P-450 expression systems, are considered. The dependence of the mouse P-450 IA1 specific activity on the cytochrome expression level in yeast microsomes is studied and results demonstrate that the low amounts of endogenous NADPH-cytochrome P-450 reductase and cytochrome b5 which are naturally present, are limiting for the heterologous monooxygenase activities. The sequences encoding human liver cytochrome b5, the native and a modified form of the yeast NADPH-cytochrome P-450 reductase were cloned by making use of PCR techniques, over-expressed in yeast as functional forms, and characterized. New vectors allowing a high level of mammalian P-450 expression upon induction were also constructed and tested. A strategy for the construction of a co-expression system allowing maximal activity of mammalian cytochrome P-450s is discussed.

Synthesis of functional mouse cytochromes P-450 P1 and chimeric P-450 P3-1 in the yeast Saccharomyces cerevisiae

Mouse liver cytochrome P-450 P1 was produced in the yeast Saccharomyces cerevisiae transformed by various expression vectors. The relative efficiency of the phosphoglycerate kinase and GAL10-CYC1 promoters to direct the P-450 P1 mRNA synthesis was determined. The level of protein synthesis was found to be dependent on the amount of the 5'-noncoding sequence of the original cDNA removed during the construction. Yeast-synthesised P-450 P1 was found to be integrated into the microsomal membrane in a fully functional form, as judged by Western blotting, optical spectra and enzymatic activities. The amount of P-450 reached up to 0.6% of the microsomal protein level. A nucleotide sequence coding for a chimeric enzyme in which 40 N-terminal codons of P-450 P1 were replaced by 36 N-terminal codons of P-450 P3 was constructed and expressed in yeast. The resulting protein retained full P-450 P1 activity and was produced with a similar efficiency suggesting that the P-450 N-terminal sequence is not involved in structures critical for the substrate specificities of the P1 isoenzyme.