An overview of the clinical characteristics of HGPS will be presented to provide a background for the workshop discussions. Approximately 80% of cases are classical in presentation, while 20% have atypical features. These atypical cases may suggest other mutations in LMNA than the G608G mutation. A comparison with other laminopathies will be made.

Mutations in LMNA, in addition to causing HGPS, have now been found to be the cause of probably more than 7 different recessive and dominant disorders, including Emery–Dreifuss muscular dystrophy types 2&3, a form of dilated cardiomyopathy, the Dunnigan type of familial partial lipodystrophy, limb girdle muscular dystrophy type 1B, Charcot–Marie–Tooth disorder type 2B1, and mandibuloacral dysplasia.

1. Emery-Dreifuss muscular dystrophy (EMD), characterized by early contractures of elbow and ankle tendons, slowly progressive muscle wasting, and a cardiomyopathy, has three modes; X-linked (EMD1, MIM 310300), autosomal dominant (EMD2, 181350), and recessive (EMD3, 604929). The X-linked EMD1 is due to mutations in the gene for Emerin. Different mutations in LMNA can cause both dominant EMD2, and recessive EMD3 forms of EMD (Raffaele 00). Lmna-/- mice show features resembling most closely EMD (Sullivan 99).

2. Mutations in LMNA underlie an autosomal dominant disease, which features dilated cardiomyopathy and conduction system disease (CMD1A, MIM 115200). Such CMD1A patients have no evidence of the joint contractures or skeletal myopathy characteristic of EDM.

3. Mutations in LMNA cause the Dunnigan type of dominant familial partial lipodystrophy (FPLD; MIM 151660). Such patients are born with normal fat distribution, but after puberty experience regional and progressive adipocyte degeneration, often associated with profound insulin resistance and diabetes.

4. Mutations in LMNA underlie the autosomal dominant LimbGirdle Muscular Dystrophy 1B (LGMD1B; MIM 159001), which is characterized by a slowly progressive pelvic girdle muscular weakness with late involvement of humeral muscles and sparing of the peroneal and tibial muscles, with age-related atrioventricular cardiac conduction disturbances and absence of early contractures.

5. Mutations in LMNA cause an autosomal recessive axonal neuropathy known as Charcot-Marie-Tooth disease 2B1 (CMT2B1; MIM 605588). The main clinical symptoms in 90% of cases are early onset, symmetrical muscle weakness and wasting (predominantly in the distal lower limbs); foot deformities and walking difficulties associated with reduced or absent tendon reflexes.

6. Mutations in LMNA underlie the autosomal recessive Mandibuloacral Dysplasia (MAD, MIM 248370) characterized by postnatal growth retardation, craniofacial anomalies, and skeletal malformations with acroosteolysis. Such patients frequently have partial lipodystrophy and insulin resistance as seen in FPLD

7. LMNA mutations also are found in several adult Werner syndrome progeric patients with atypical features that were negative for WRN mutations. These have now been found positive for LMNA mutations other than G608G (J. Oshima).

8. A SNP in LMNA is associated with obesity traits in Native Eskimo (Inuit) populations (Hegele 01).

9. A mouse model of HGPS has been reported with homozygous mutations in LMNA at L530P, which when heterozygous in humans causes AD-EMD (Mounkes 03).

Mutations in Lamin A are the Cause of Hutchinson-Gilford Progeria Syndrome

Francis S. Collins, M.D., Ph.D., Maria Eriksson, Ph.D.
Genome Technology Branch, NHGRI

Positional cloning of Mendelian disorders has now succeeded in identifying the responsible gene for over 1,600 such conditions. However, the strategy has almost invariably involved the study of families with multiple affected individuals. Hutchinson-Gilford Progeria Syndrome (HGPS) presented a particular challenge because of the absence of well-documented instances of recurrence of the classic form of the disease. We undertook a genome wide scan of DNA samples from 23 classic cases, searching for any evidence of homozygosity that might indicate the location of a rare recessive allele. To our surprise, we identified 3 cases with abnormalities of the long arm of chromosome 1, including two with uniparental isodisomy and one with a paternal 6 Mb deletion. Of the 80 genes in this candidate interval, lamin A/C (LMNA) emerged as a highly attractive candidate, given its role in several other genetic disorders. Sequencing of DNA from affected individuals revealed that 18 out of 20 of these harbored a point mutation in codon 608 (GGC to GGT) of exon 11. A single patient had a different mutation (GGC to AGC) in codon 608, and another single case had a mutation in exon 2 (E145K).

The common point mutation was not found in any of the unaffected parents, indicating that it arose de novo. We were surprised, however, to note that it did not change an amino acid. Subsequent investigation demonstrated that this mutation (G608G) creates an abnormal splice donor, producing an RNA which deletes 150 nucleotides from the coding region of the RNA, resulting in a protein with 50 amino acids missing very near the C-terminus. Examination of nuclei from fibroblasts of progeria patients using antibodies against lamin A demonstrates dramatic structural abnormalities, with nuclear blebbing in a significant fraction of the cells.

We hypothesized that the abnormal protein acts as dominant negative to disrupt the normal structure of the nuclear lamina. To further investigate this phenomenon, we have undertaken the construction of transgenic mice, engineered to carry the common exon 11 mutation. Beginning with a bacterial artificial chromosome (BAC) containing the entire LMNA gene on a 164 kb genomic segment, we have reengineered the BAC to contain the G608G mutation. That construct has now been injected into transgenic mice, both as a circular and a linear construct, and transgene positive animals have just been born. A second type of construct involves the creation of “minigene” carrying the G608G mutation, driven by a tetracycline-inducible promoter. A third construct utilizes a BAC that has been reengineered to allow tissue-specific switching from the wild type to mutant sequence, using the cre-lox system.

LMNA Mutations in Progeroid Syndromes

Lishan Chen, Nancy B. Hanson, George M. Martin, Junko Oshima
Department of Pathology, University of Washington, Seattle, WA

Approximately 80-90% of cases with a clinical diagnosis of Werner syndrome (WS, “Progeria of Adults”) have demonstrable mutations in the WRN gene. We recently found that mutant forms of LMNA, distinct from what has been observed in HGPS, are responsible for a subset of these “atypical” forms of WS, in which WRN is not mutated (4). Two identified mutations, R133L and L140R, along with E145K, are located in the heptad repeat region of a- coiled coil domain segment 1B. The heptad region is unique to nuclear intermediate filaments and not present in cytoplasmic ones.

We have begun to ask whether the cellular phenotypes caused by the LMNA mutations can be reversed by short interfering RNA (siRNA) targeted to the mutant LMNA mRNA. We introduced expression vectors coding for 21-mer dsRNAs by lentiviral system into the HG fibroblasts or atypical WS fibroblasts immortalized with hTERT. Characterization of these lines are currently in progress.

Laminopathy Models in Mice

Colin L. Stewart
Mammalian Development Section, CDBL, National Cancer Institute FCRDC, Frederick MD 21702

The lamins are nuclear intermediate filaments that are major determinants of nuclear morphology and are thought to be important in regulating chromatin organization and DNA replication. Over the past 3 years, at least 7 diseases, the laminopathies, have been linked to different mutations in the A-type Lamins. These can be largely grouped in to 2 classes, the striated muscle laminopathies that affect the heart and skeletal muscle. These include Autosomal dominant Emery- Dreifuss Muscular Dystrophy, (AD-EDMD), Dilated Cardiomyopathy (DCM) and LimbGirdle Muscular Dystrophy (LGMD-1B) and possibly Charcot-Marie-Tooth (CMT2B), a peripheral neuropathy leading to muscle wasting. The other class is the lipodystrophies which include Familial Partial lipodystrophy (FPLD) and Mandibuloacral Disease (MAD). Recently a 7th laminopathy has been described, Hutchinson-Gilford Progeria (HGPS). Children afflicted with HGPS, are growth retarded, have a multitude of abnormalities affecting the skin and skeletal system and usually die in their teens from cardiovascular problems. We have derived novel mouse lines carrying a variety of the mutations associated with the different laminopathies. Among these is a mouse model for HGPS, which was introduced as a splicing defect in one of the 3’ exons. Mice carrying this mutation show postnatal growth retardation, hypoplasia in the skeletal and cardiac tissues, thin, parchment like skin, and bone defects. The mice die prematurely at 4-5 weeks. The mouse lines are recapitulating the phenotypes associated with the different laminopathies and are providing novel insights into the molecular and cellular basis of these diseases.

Lamin A-structure, function, and protein interactions in the normal and defective states

Howard J. Worman

The nuclear lamina is a meshwork of intermediate filaments primarily underlying the inner aspect of the inner nuclear membrane that also extends into the nuclear interior. The lamina is associated with the inner nuclear membrane as a result of the interactions between its protein building blocks, the lamins, and integral proteins of this membrane, including LBR, laminaassociated polypeptides and emerin. Lamins and some of the integral proteins of the inner nuclear membrane also interact with chromatin. Lamin filaments are also connected to the inner aspects of the nuclear pore complexes. Because of its interactions with the chromatin and pore complexes, the lamina has been proposed to function in the regulation of gene expression. However, the precise function of the nuclear lamina in regulating gene expression is not understood.

In humans, 3 genetic loci encode several different lamins that polymerize to form the lamina. Lamins A and lamin C arise as a result of alternative splicing of RNA encoded by LMNA, a gene localized to chromosome 1q21.2. Lamins A and C are expressed in most but not all somatic cells. They are identical for their first 566 amino acids. Lamin C has 6 unique carboxyl-terminal amino acids. Prelamin A, a precursor of lamin A, has 98 unique amino acids and is farnesylated at its carboxyl-terminus after synthesis. The last 18 amino acids, containing the farnesyl group, are removed by endoproteolytic cleavage. Lamins A and C, like all intermediate filament proteins, have a head domain, highly conserved alpha-helical rod domain and carboxyl-terminal tail domains. Nuclear localization signals in the tail domain signal targeting of lamins to the nucleus. The common portion of the tail domain of lamins A and C contain an immunoglobulinlike fold, a domain found in many other proteins that often mediates protein-protein and proteinnucleic acid interactions.

Mutations in nuclear lamins A and C have been shown to cause several different diseases including autosomal dominant Emery-Dreifuss muscular dystrophy, Dunnigan-type partial lipodystrophy, Charcot-Marie-Tooth disorder type 2B1, mandibuloacral dysplasia and Hutchinson-Gilford progeria. Some lamin A mutants that cause muscular dystrophy appear to be stably expressed in cells and lead to abnormal nuclear structure visible by light microscopy. In muscular dystrophy, most of the mutations are missense mutations or small mutations scattered throughout the molecules. In Dunnigan-type partial lipodystrophy, the vast majority of mutations change the charge character of a solvent-exposed surface of the immunoglobuliln-like fold, suggesting that a specific interaction may be altered. In contrast, mutations in the immunoglobulin-like fold in muscle diseases lead to an overall disruption of structure. The LMNA mutation that causes Hutchison-Gilford progeria leads to loss of 50 amino acids from the carboxyl-terminal tail of lamin A. It is not yet clear how this mutant lamin A, which may be expressed a low levels, alters nuclear function.

A deficiency in Zmpste24, a prelamin A-processing enzyme, causes spontaneous bone fractures and muscle weakness.

Stephen G. Young†, Susan Michaelis‡, and Martin O. Bergo†. †Gladstone Institute of Cardiovascular Disease, University of California, San Francisco, CA 94141-9100 and ‡Department of Cell Biology, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205.

Zmpste24, an integral membrane metalloproteinase of the endoplasmic reticulum, has been implicated in the posttranslational processing of prenylated “CAAX” proteins. Zmpste24 is the murine orthologue of yeast Ste24p, an endoprotease that is required for the biogenesis of the mating pheromone a-factor (a farnesylated CAAX protein). In yeast, Ste24p actually has dual roles in the endoproteolytic processing of the a-factor precursor—cleaving three amino acids from the carboxyl terminus of the protein (the –AAX of the “CAAX motif”) and cleaving seven amino acids from the amino terminus of the protein. Remarkably, mouse Zmpste24 is quite capable of carrying out both of these “a-factor endoproteolytic processing reactions,” despite the fact that no orthologue for a-factor been identified in mammals. Moreover, Zmpste24 is the only mammalian protease that is capable of carrying out the amino-terminal endoproteolytic processing of yeast a-factor. Membranes from cells and tissues of Zmpste24-deficient mice are incapable of carrying out the amino-terminal processing of a-factor, whereas membranes from the cells of wild-type mice faithfully carry out this reaction.

Prelamin A—the precursor to lamin A—is a farnesylated CAAX protein in mammals that undergoes more than one endoproteolytic processing reaction. Prelamin A is cleaved once to release the C-terminal –AAX, and then a second time to release an additional 15 residues from the C terminus. Because prelamin A in mammals—like a-factor in yeast—undergoes more than one endoproteolytic processing reaction, we hypothesized that Zmpste24 might play a role in prelamin A processing. Indeed, this proved to be the case. Absolutely no mature lamin A is formed in Zmpste24-deficient cells, and an incompletely processed prelamin A (a prelamin A that has not undergone the second proteolytic processing reaction) accumulates in the cells. Our data suggest that Zmpste24 is required for the occurrence of the second prelamin A endoproteolytic processing reaction. We have hypothesized that Zmpste24 could participate in the cleavage of the –AAX from the C-terminus of prelamin A, but this has not yet been proven. Zmpste24–/– mice appear virtually normal at birth, but go on to manifest several phenotypes that are suggestive of laminopathies and/or progeria. They gain weight slowly; they develop micrognathia, dental abnormalities, and progressive hair loss. They develop osteolysis in multiple bones, leading to spontaneous bone fractures. Cortical and trabecular bone volumes are significantly reduced, compared with wild-type mice. Zmpste24–/– mice also manifest muscle weakness in the lower and upper extremities, resembling mice lacking lamin A/C. However, the etiology of the muscle weakness phenotype in the Zmpste24–/– mice is not entirely clear. In contrast to the lamin A/C-deficient mice, the Zmpste24–/– mice did not have obvious dystrophic changes in skeletal muscle.

The nuclei of Zmpste24-deficient cells are misshapen. We suspect that both the muscle weakness and the bone abnormalities in Zmpste24–/– mice are due to defective biogenesis of lamin A. In the future, it will be interesting to compare Zmpste24-deficient mice with authentic mouse models of Hutchinson-Gilford progeria syndrome.

Can ZmpSte24 be Prompted to Process the HGPS Form of Prelamin A?
Insights from Saccharomyces cerevisiae a-factor Processing

Susan Michaelis, Gregory Huyer, Monica Mallampalli, Meredith Boyle, Franklin Nouvet, Amy Kistler, Walter K. Schmidt, Amy Tam, and Konomi Fujimura-Kamada

Many cellular proteins are synthesized as precursors that undergo elaborate post-translational processing events critical to their normal function. To better understand these events, we are studying the biogenesis of the Saccharomyces cerevisiae mating pheromone a-factor, an extracellular signaling peptide that is prenylated and carboxyl methylated. The biogenesis of the a-factor precursor is distinctive, involving: (1) C-terminal (CaaX) processing (by a farnesyl transferase, the endoproteases Ste24p and Rce1p, and the methyltransferase Ste14p/ICMT), (2) N-terminal proteolytic cleavage (by the endoproteases Ste24p and Axl1p/IDE), and (3) a nonclassical export mechanism (mediated by the ABC transporter Ste6p). We have recently shown that one of the a-factor processing components, Ste24p, is a multispanning membrane protein with intrinsic zinc metalloprotease protease activity that acts in dual steps of a-factor biogenesis, mediating both C- and N-terminal proteolytic processing of the a-factor precursor (1, 2, 3). In extending our studies to mammalian systems we have demonstrated that the mammalian zinc metalloprotease homologue, Zmpste24, can substitute for yeast Ste24p in a-factor biogenesis (2).

Mammalian prelamin A, like the yeast a-factor precursor, undergoes dual processing (Cterminal CaaX endoproteolysis, followed by a nearby “internal” proteolytic cleavage), prompting us to propose that prelamin A processing might rely on Zmpste24. In a collaborative study with Dr. Stephen G. Young’s group (4) we showed that prelamin A processing is indeed Ste24p-dependent (although it should be noted that the precise proteolytic site(s) cleaved by Zmpste24 remain to be definitively demonstrated). The Zmpste24-/- knockout mouse phenotype exhibits striking similarities with human laminopathies, most closely resembling mandibuloacryl dysplasia (MAD). Furthermore, a recent study indicates that while some individuals with MAD have mutations that map in the lamin A gene, others have mutations that map in Zmpste24 (5). Interestingly, cell lines derived from HGPS individuals produce a shortened form of prelamin A that lacks the internal proteolytic cleavage site. We hypothesize that a lack of correct prelamin A processing may lead to at least some of the pathological phenotypes observed in HGPS, possibly because the persistence of prenylated unprocessed (or misprocessed) mutant prelamin A is detrimental. If this were the case, then restored processing at a novel site within prelamin A in HGPS cells could potentially alleviate some of the phenotypes of HGPS. We will present evidence from our studies of yeast a-factor processing that suggest that Ste24p cleavage may exhibit some “leeway” (as well as some specificity) in terms of amino acid recognition, which could potentially have implications for seeking pharmacological intervention for HGPS.

References

1. Fujimura-Kamada, K., Nouvet, F. J., and Michaelis, S. (1997) J. Cell Biol. 136:271-85.
2. Tam, A., et al. (1998) J. Cell Biol. 142:635-49.
3. Tam, A., Schmidt, W. K., and Michaelis, S. (2001) J. Biol. Chem. 276:46798-806
4. Bergo, M. O. et al. (2002) Proc. Natl. Acad. Sci. USA 99:13049-54
5. Agarwal, A. et al. (2003), Hum. Mol. Biol. (In Press)

Aging Phenotypes and Cellular Responses

Judith Campisi

There are now a number of examples of single gene mutations that cause segmental accelerated aging syndromes in humans, as well as mammalian model organisms such as mice. One major gap in our knowledge and ability to treat these progeroid syndromes is our relatively poor understanding of how the causative mutations lead ultimately to the premature aging and other characteristic phenotypes of the organism.

I will discuss the idea that, in many cases, substantial progress can be made in closing this gap by studying the responses and phenotypes of cells cultured from individuals carrying "premature aging" mutations. I will describe a system we recently developed for reversibly expanding certain mutant human cell populations, which we believe can greatly facilitate such studies. As an example of this approach, I will briefly describe some of our recent studies using cells from individuals with hereditary defects in genes related to that responsible for the Werner syndrome. Werner syndrome is perhaps the best studied of the premature aging syndromes in humans, and is caused by inactivating mutations in WRN, a DNA exonuclease/helicase and member of a family of related helicases that very likely participate in one or more aspect of genomic maintenance. I will describe how studies using cells from individuals with Bloom syndrome, an early onset cancer prone syndrome caused by inactivating mutations in BLM, a DNA helicase related to WRN, have provided insights into two prominent organismal phenotypes associated with Bloom syndrome. These phenotypes are the severe pre- and postnatal growth retardation and the extremely high incidence of cancer that Bloom syndrome individuals develop in the first or second decades of life. I will also describe some recent studies using cells from individuals with Werner syndrome which suggest that structural components of the nucleus may play a role in mediating some functions of WRN. These findings raise the possibility that there may be some overlap in one or more process that is defective in the Hutchinson-Gilford progeria syndrome and the Werner syndrome.

Nuclear Lamin Structure and Function In Healthy and Diseased Cells

Robert D. Goldman, Department of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, Chicago

The nuclear lamins are Type V intermediate filament (IF) proteins that are thought to be the evolutionary progenitors of cytoplasmic IF proteins such as keratin. Humans possess three lamin genes (LMNA, LMNB1, and LMNB2), which yield four A-type (LA, LA D 10, LC, LC2), and three B-type lamins (LB1, LB2, LB3) due to alternative splicing. All vertebrate cells express at least one lamin B (LB), whereas A/C-type lamins (LA) are developmentally regulated and are expressed primarily in differentiated cells. This suggests that lamins have different functions related to cell and tissue type specificity and that they are somehow involved in normal development. Lamins contain three distinct domains, required for their normal assembly. The a- helical central rod domain, consisting mainly of heptad repeats, separates the non-a-helical Nand C-terminal domains. A highly conserved portion of the C-terminal domain forms an immunoglobulin-like fold. The central rod domain is essential for the formation of the coiledcoils responsible for the assembly of lamin dimers, the basic building blocks of the polymerized structure that forms the major component of the nuclear lamina. During interphase, this lamina forms a molecular interface between the inner nuclear envelope membrane and chromatin. Although the structure of the lamins in the lamina are poorly defined for most nuclei, it has been shown that they are involved in determining the shape and mechanical stability of the nucleus during interphase and the disassembly/reassembly of the nucleus during cell division. These proteins are also distributed throughout the nucleoplasm where they are involved in DNA replication, transcription, chromatin and nuclear pore organization, as well as the organization of various lamin associated proteins (LAPs) including inner nuclear envelope membrane components. The importance of the nuclear lamins is reflected in an ever-increasing number of human diseases linked to mutations in the human lamin A (hLA) gene (LMNA). Collectively these diseases, frequently displaying overlapping phenotypes, are called “laminopathies”. They include various forms of muscular dystrophy, cardiomyopathies, lipodystrophies and most recently the premature aging disease, Hutchinson-Gilford Progeria Syndrome (HGPS). To date, a detailed search of the literature reveals that there are over 80 mutations in human LMNA that cause disease and these are distributed over all regions of LMNA with the exception of exon 10. Interestingly, many cells from patients with laminopathies (in addition to cells from the LMNA - /- mouse) have abnormally shaped nuclei with weakened areas termed blebs. Our preliminary studies of cells obtained from Progeria patients have revealed gross abnormalities in nuclear architecture both at the light and electron microscopic level, including a thickening of the lamina, and an abnormal distribution of nuclear pores. These findings will be discussed in light of our hypothesis that nucleoplasmic lamin structures, in addition to those in the lamina, form a nucleoskeletal system that provides the infrastructure required for numerous nuclear functions, including DNA replication and transcription. Understanding how these functions are altered by HGPS mutations will shed light on the mechanisms responsible for the multiple age-related disorders seen in patients with progeria, including cardiomyopathies and strokes. Supported by a grant from the National Cancer Institute.

Serum Lipid and C-reactive Protein levels in Hutchinson-Gilford Progeria Syndrome

Leslie B. Gordon, Ingrid A. Harten, Susan M. Jalbert and Alice H. Lichtenstein

Individuals with Hutchinson-Gilford Progeria Syndrome (HGPS) develop premature severe atherosclerosis in childhood and frequently succumb to the disease before the second decade of life. Despite this observation strikingly little is know about plasma lipid and lipoprotein patterns in children with HGPS. In addition, C-reactive protein (CRP), a new and lipid-independent risk factor for cardiovascular disease, has never been assessed in children with HGPS.

We have analyzed serum lipid profiles of 16 children with HGPS (insert age range of children) and compared them with age and sex-matched controls. Total, HDL, and LDL cholesterol levels were not significantly different between the two groups of children (158.0 ±30.9, 35.4 ±15.1, 88.9 ±24.9 and 153.3 ±22.0, 41.3 ±10.6, 80.8 ±20.3mg/dL, HGPS and control subjects, respectively). Although statistically insignificant, LDL/HDL ratios inHGPS children (3.1 ±1.8) are mildly elevated compared to the control children (2.1 ±0.9) (P=0.062).

We have also analyzed CRP levels in HGPS serum samples, and compared them with age-matched controls. CRP is a general marker for inflammation that has recently been implicated in adults as a risk factor for genetic predisposition towards cardiovascular disease. CRP levels were similar between HGPS and control children (1.3 ± 2.3 and 2.9±7.0 mg/dL, respectively).

Thus, the pathobiological basis of cardiovascular disease in HGPS remains to be determined. Differences in the atherogenic properties of the LDL particle, non-functional HDL particles, or differences upstream of these mediators of atherosclerotic plaque formation cannot be ruled out.

Discussion points:

Should children with HGPS be placed on statins? Should statin treatment depend on lipid profiles?

What does this data imply with respect to similarity in development of atherosclerosis between the normal aging adult and children with Progeria? Taking into consideration the limited amount of serum available are there other measures that should be made in order the answer the aforementioned questions?

Of Mice and Men: Atherosclerosis and Aging

Stephen M. Schwartz, M.D., Ph.D

Atherosclerosis is a disease that has no clinical outcome until humans are well past the age of procreation. Studies of the role of aging in atherosclerosis may depend on understanding the time course of the critical events in the natural history of this disease in mice and men.

The earliest event in man seems to be focal accumulations of “intimal smooth muscle cells.” at lesion prone sites in normal arteries. These structures accumulate lipid in what may be an irreversible fashion. The other early event, requiring hyperlipidemia in mice, is the accumulation of lipid in intimal macrophages to form “fatty streaks.” Unlike intimal cell masses, fatty streaks are reversible. If the fatty streak lasts long enough it acquires smooth muscle and becomes irreversible. Evidence that intimal cell masses are the precursors of the adult human atherosclerotic lesion includes the monoclonality of the smooth muscle cells that make up the adult lesion. We have shown that these cells have a unique transcriptional signature. This and other evidence, suggests these cells may have an origin different from that of medial smooth muscle. The formation of a fibrous cap of intimal smooth muscle over a fatty streak and the subsequent necrosis of the macrophage to form a fatty core complete the process of creation of a classical lesion as described 150 years ago by Virchow.

Over 3-4 decades in man, or about 1 year in mice, continued hyperlipidemia converts these early lesions into clinically significant lesions with a central necrotic core and plaque rupture. This process takes almost a year of the mouse’s 2-3 yr. life span. The lack of selective pressure in these late events may be illustrated by the “metabolic syndrome” of diabetes, hyperlipidemia, and hypertension. Geneticists believe this syndrome conveyed a metabolic advantage to our ancestors, but today causes “premature” death in 50 and 60 year olds.

Almost all death from atherosclerosis is the result of arterial occlusion arising from thrombosis/ coagulation. Elevation of systemic coagulative states in man accelerates death from atherosclerosis. The issue for aging, however, is probably the development of a pro-coagulant, prothrombotic, rupture prone lesion. We lack a murine model for this process but, in man, the conventional wisdom is that coagulation results from plaque rupture with exposure of pro-coagulant proteins that accumulate as the plaque progresses. Surprisingly, however, studies of the ability of plaque to initiate the extrinsic clotting pathway ignore the fact that even without plaque rupture, the lesions contain abundant coagulation factors. In contrast, a common form of occlusion in 40 and 50 year old humans, called “erosion” by Virmani, occurs in the apparent absence of intimal clotting factors. The explanation may be the recent discovery by Nemerson of a circulating form of tissue factor that becomes activates during vasoocclusion. The pharmacology of vasospasm remains unknown although we know that a very early event in atherosclerosis is the loss of the normal vaso-protective effects of NO. Other processes that contribute to vasospasm may be the accumulation of oxidation products as lesions progress.

The final events in plaque progression are scarring and rupture. Both events occur in the murine model. In both species the scarring leads to contraction of the vessel wall and loss of lumen. Otherwise, we can not be sure the processes in the mouse are the same as in humans. For example, there is no evidence of intimal cell masses in the mouse and rupture of the murine lesions does not seem to lead to coagulation. Moreover, in our mice, rupture is attributable to death of macrophage in the plaque, but in humans, the evidence suggests that the intimal smooth muscle cap becomes atrophic, perhaps through accelerated apoptosis, and the matrix is digested by enzymes released from inflammatory cells. In this case, however, the proteases in early murine lesions do not seem to cause these lesions to rupture, even when overexpressed, implying that we need to know more about the molecules determining plaque rupture, even in the mouse.

Skeletal Findings in Hutchinson Gilford Progeria Syndrome – What? Where? Why?

Heather S. Hardie, MD

HGPS is known to cause many reproducible abnormalities in the skeletal system including distal clavicular and tuftal resorption, coxa valga deformity of the hips, delayed closure of fontanelles, and microagnathia. In the past, it has been assumed that many of the skeletal features are part of the “premature aging” process.

The skeletal findings will be presented to show that they are, in fact, caused by a skeletal dysplasia. We will review basic bone formation and growth in order to demonstrate possible patterns of bone involvement in HGPS. It appears that HGPS primarily affects membrane bones and may disproportionately affect bones undergoing intramembranous ossification and/or sites of secondary bone formation.

We will also review the processes of and relationship between osteoblastogenesis and osteoclastogenesis as they are now understood on the molecular level to try to elucidate any possible connection to HGPS. Known molecular defects in related conditions and conditions with related findings will be discussed hoping to determine any similarities to HGPS.

MicroarrayStudies of mRNA levels in HGPS

¹Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, Providence, RI 02912, USA. ²Children's Hospital Informatics Program, Boston, Massachusetts 02115, USA., *workshop presenter.

Supported in part by a grant from the Progeria Research Foundation.

Aggrecan expression is substantially upregulated in Hutchinson-Gilford progeria syndrome skin fibroblasts

Quantitative real time RT-PCR confirmed a previously reported elevation of platelet-derived growth factor A-chain (PDGF-A) in progeria [Winkles, J.A., O’Connor, M.L. & Friesel, R. (1990) J. Cell. Physiol. 144: 313-25] with transcript levels 13 ± 2 times higher in AG03513B fibroblasts and 11 ± 2 times higher in AG03198A fibroblasts than in their respective agematched controls. In addition, two new genes were identified as having significantly altered transcript levels. Substantially increased and decreased aggrecan and nucleotide pyrophosphatase transcript levels were seen, respectively. Aggrecan mRNA was extremely elevated in two progeria strains (24 ± 5 times and 41 ± 4 times that of chronologically age-matched controls in AG03198 and AG03513B respectively). In contrast, nucleotide pyrophosphatase transcript levels were substantially decreased, with a differential expression of –13 ± 3 in AG03198 fibroblasts and –59 ± 3 in AG03513 fibroblasts. The aggrecan transcript is a promising candidate for use as a fibroblast marker for the syndrome. Collaborative analysis with Dr. Joan Lemire, Tufts, appears to show an unquantified dramatic increase in aggrecan in the progeria line HGADFN001. Two blots were duplicate panels run on a single gel and transferred together. This is currently being pursued with further cell lines and with Dr. Amanda Fosang, Melbourne in quantitative studies. Aggrecan is a major proteoglycan component of cartilaginous tissue but is not normally expressed at appreciable levels in fibroblasts. We propose a model of poorly controlled proteoglycan expression and secretion in progeria fibroblasts. Elevated aggrecan is known to occur in osteoarthritis and cartilage remodeling, typically during destruction and inflammation. Uncontrolled proteoglycan expression would be expected to have profound effects on the properties and cellular responses of progeria connective tissue.

NA inhibition applications to HGPS

Dr. Gregory Hannon

The use of biochemical systems from Drosophilia and genetic studies in plants and invertebraes have begun to reveal a mechanistic basis for RNA interference and related phenomena. The canonical model involves a two-step mechanism. We identified an RNAseIII family nuclease, Dicer, which initiates RNAi by processing dsRNA silencing triggers into small RNAs of ~22 nt in length. These enter an effector complex RISC, which seeks out and degrades homologous subtstrates. Genetic studies of Dicer-null animals (i.e., C. elegans) have suggested roles for the RNAi machinery in the regulation of endogenous genes. Specifically, Dicer and components of the RISC complex have been implicated in processing of and in gene regulation by endogenously encoded small hairpin RNAs, known collectively as microRNAs (miRNAs). We exploited these observations to test the possibility that miRNAs might be remodeled to regulate genes of interest.

We, and others, have shown that expression of shRNAs from RNA polymerase III or RNA polymerase II promotes results in silencing of homologous genes. We have recently extended these findings to living animals. We continue to pursue parallel paths toward a deeper understanding of the underlying mechanism of RNAi and toward expanding the applications of RNAi as a tool for investigating gene function in mammals.

Drug screening and development for Progeria

Christopher P. Austin, M.D.

The discovery of Lamin A gene mutations and a characteristic nuclear phenotype in HGPS opens the possibility for screening for small molecules that might reverse the cellular, and eventually the clinical, manifestations of the disease. Though greater understanding of the cell biological effects of HGPS-associated Lamin A mutations will facilitate effective drug screening, general principles of assay development, compound screening, lead identification, and drug development can be discussed in the context of what is currently known about HGPS. This presentation will cover these principles with a view toward what research will be most helpful in designing effective and physiologically relevant drug screens. Plans currently underway to initiate largescale compound screening at NIH will also be discussed.

Vascular smooth muscle cell proliferation, cell cycle, and bone marrow-derived progenitor cells: possible relevance to HGPS

Vascular regeneration and repair are essential to the survival of adult blood vessels. Arterial wound repair requires the coordinated temporal and spatial expression of proteins that regulate vascular cell proliferation. p27^Kip1 is an important inhibitor of cell division. A member of the Cip/Kip family of cyclin-dependent kinase inhibitors (CKIs), p27^Kip1 binds and alters the activities of cyclin D-, cyclin E- and cyclin A-dependent kinases in quiescent cells, leading to failure of G1/S transition and cell cycle arrest. An increase in the levels of p27^Kip1 causes proliferating cells to exit the cell cycle, and a decrease in p27Kip1 is necessary for quiescent cells to resume division.

Bone marrow (BM)-derived progenitor cells are increasingly recognized as key components in vascular regeneration. After vascular injury, progenitor cells from arterial and BM compartments are mobilized by cytokine activation. BM-derived progenitor cells travel through the circulation and home at sites of vascular damage, proliferate and form an arterial lesion, along with cellular components from the native artery. This process is not well understood, and the molecular regulators are mostly unknown.

We find that that p27^Kip1 plays a major role in the regulation of cardiovascular disease through its effects on vascular regeneration and repair. p27^Kip1 directly regulates vascular remodeling through modulation of BM-derived progenitor cell function, vascular smooth muscle cell (vsmc) proliferation, and inflammation. In a mouse model of mechanical arterial injury, homozygous deletion of p27^Kip1 produces accelerated proliferation of vsmcs, a large arterial neointima that occludes the vessel lumen, and intense vascular inflammation (neutrophils, monocytes/macrophages, T-lymphocytes). To understand the origins of vsmc and inflammatory cells, bone marrow transplantation (BMT) studies were performed. We find that formation of an arterial lesion is plastic; that is, BM-derived progenitor cells substantially repopulate the vascular lesion when p27-/- BM is transplanted into p27+/+ mice subjected to mechanical injury while arterial lesions from p27+/+ BM transplanted into p27-/- mice consist primarily of vsmcs from the native artery. The mechanism accounting for the p27-/- selective advantage is an increase in BM progenitor pool size in p27-/- mice. In addition, we tested whether the absence of p27Kip1 permits the clonal expansion of T- and B-lymphocytes in BMT studies between RAG-/- and p27-/- mice. While RAG-/- mice have reduced inflammation, p27-/- BM produced equally large vsmc-rich proliferative arterial lesions in RAG-/- and RAG+/+, providing again direct evidence for p27Kip1 regulation of proliferation in vsmc and other cell lineages in CV disease. In summary, these findings suggest that repair and regeneration in the vasculature is a dynamic process mediated by in part by BM-derived stem cells and cell cycle regulation of proliferation and inflammation.

Darwin J. Prockop, M.D., Ph.D.