| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Schulich School of Medicine and Dentistry University of Western Ontario London, Ontario, Canada N6A 5K8
Address all correspondence and requests for reprints to: Robert A. Hegele, M.D., FRCP(C), FACP, Blackburn Cardiovascular Genetics Laboratory, Robarts Research Institute, 406-100 Perth Drive, London, Ontario, Canada N6A 5K8. E-mail: hegele{at}robarts.ca.
Characterizing the molecular genetics and defining the clinical phenotypes of the lamin A/C-related disorders, collectively referred to as "laminopathies," has been one of the most intriguing chapters in human genetics (1). Lamins are intermediate filament proteins that reside normally on the inside of the nuclear envelope and perform numerous biological functions including maintenance of structural integrity, organization of chromatin, governing division of nuclear contents during mitosis and meiosis, and regulation of access of transcription factors to the nucleus (2). In the last 10 yr, mutations in the genes encoding these proteins have been shown to underlie a breathtaking array of human illnesses, including inherited muscular dystrophies, peripheral neuropathies, cardiomyopathies with conduction disturbances, and disorders of accelerated aging (1).
However, for endocrinologists, the subgroup of laminopathies that has held the greatest fascination has been those with a primary metabolic phenotype, specifically Dunnigan-type familial partial lipodystrophy subtype 2 (FPLD2; MIM 151660), which results from heterozygous mutations in LMNA encoding lamin A/C (3). The partial lipodystrophy syndromes are associated with insulin resistance, multiple quantitative metabolic disturbances, and ultimately diabetes. FPLD2, which although very rare is still probably the most prevalent of the inherited lipodystrophies, is characterized by sc lipoatrophy of arms, legs, gluteal region, and abdomen, with lipohypertrophy of visceral fat stores, upper truncal fat, and neck and facial fat (4).
The identification of LMNA as the causative gene for FPLD2 required classical linkage analysis studies which showed that genetic markers on chromosome 1q21 perfectly segregated with the FPLD2 phenotype (5). This statistically implicated this region of the genome as harboring the causative gene for FPLD2 (5). Subsequent detailed analysis of genes residing within 1q21 was performed using automated sequencing of genomic DNA of members of a Canadian FPLD2 family. These studies revealed a novel heterozygous missense mutation in exon 8 of LMNA, namely p.R482Q (6). The list of causative mutations in FPLD2 was rapidly expanded soon thereafter, but the overwhelming majority of FPLD2 mutations were found in LMNA exon 8, especially codon 482, with missense mutations p.R482Q, p.R482W, and p.R482L (7). Presently, most newly diagnosed North American patients with FPLD2 have one of these codon 482 mutations. Virtually all FPLD2 subjects with codon 482 mutations had a pure "classical" FPLD2 phenotype, with minimal involvement of other tissues, such as skeletal or cardiac muscle. We have since noted that the FPLD2 phenotype seemed milder in carriers from pedigree branches that were remote from the proband and her immediate nuclear family.
Partial lipodystrophies have been closely scrutinized using a wide range of technologies. A key approach has been "phenomic" analysis of molecularly characterized individuals. Phenomics has been defined as systematic application of clinical, biochemical, and imaging methodologies—tools that are familiar to practicing clinicians and clinician investigators (8). But while a practicing clinician might have incentives to minimize the amount of technology expended to make the correct diagnosis, a human genetics investigator, even with constrained research funding, sometimes has the luxury to apply diagnostic technologies more extensively, albeit strategically, to identify subtle clinical phenotypic differences. These "deep phenotypes" are often quantitative, so that differences between individuals with well-characterized genotypes can be evaluated statistically.
Characterizing phenotype-genotype correlations in patients can be more technically challenging, resource-intensive, and artifact-prone than studying them in animal models. But there are also advantages when studying phenotypic consequences of mutations in patients, including the fact that the landscape of background genetic and environmental factors—compared with, say, transgenic mouse models—can be more appropriate when attempting to translate experimental findings of potential relevance to human disease. For instance, careful phenomic evaluation of FPLD2 kindreds showed the initial metabolic insult to be insulin resistance followed closely by dyslipidemia, then hypertension, then diabetes, and finally vascular disease (3). The clarified progression of abnormal phenotypes in FPLD suggested a rational, staged treatment regimen to possibly modulate the natural history and development of disease end points (3). Furthermore, understanding the stages of disease progression might not only have value for rare FPLD2 patients, but could also help in understanding the metabolic evolution of common insulin resistance or metabolic syndrome (MetS).
More recent investigations of LMNA and lipodystrophy have followed three experimental paths. First, powerful molecular genetic technologies have continued to define new laminopathies resulting from new LMNA mutations. Some recently characterized multisystem laminopathies, such as Hutchinson-Gilford progeria syndrome, atypical progeroid syndromes, and mandibuloacral dysplasia, also had partial lipodystrophy as a component phenotype (1). In retrospect, it is not surprising that certain LMNA mutations lead to complex syndromes that include partial lipodystrophy as one of the elements.
Second, genomic DNA sequencing experiments expanded to allow evaluation of other candidate genes, particularly in patients with a suggestive phenotype but no germline LMNA mutations. Such experiments uncovered heterozygous dominant negative and haploinsufficiency mutations in PPARG (3, 4) encoding peroxisome proliferator-activated receptor 
. The resulting phenotype was called FPLD3 (MIM 604367) and was characterized by prominent insulin resistance but less dramatic lipoatrophy of extremities, concurrent with preservation of abdominal sc fat and relatively less visceral adiposity compared with FPLD2 (3). The AKT2 gene has also been implicated in some cases (9).
Finally, with reduced barriers to automated DNA sequencing, the net for mutation detection is being cast much more widely. Genomic interrogation of LMNA has been extended to include large numbers of subjects without classical FPLD2, but instead with more subtle—even clinically undetectable—partial lipodystrophy together with nonclassical adipose tissue involvement or with overlapping syndromes involving multiple organs and tissues. For instance, in the current issue of JCEM, Decaudain et al. (10) have extended the range and characterization of LMNA-associated metabolic phenotypes by screening the genomic DNA of 277 unrelated adults with a range of phenotypes, including lipodystrophy and/or android body habitus, insulin resistance, or altered glucose tolerance. They found LMNA mutations in only approximately 10% of patients, indicating molecular heterogeneity of the complex "relaxed" phenotype. They found typical FPLD2 LMNA codon 482 mutations in 17 patients (6.2%). However, in 10 patients (3.6%) they found a total of nine additional heterozygous LMNA non-codon 482 mutations: eight of these were missense mutations (two in exon 1, five in exon 7, and one in exon 9), and one was a frameshift mutation (in exon 11). Ex vivo experiments provided evidence of dysfunction for at least some of the mutations. They then grouped patients into those with and without codon 482 mutations. Their comprehensive phenomic analysis has led to some new insights.
First, they noted that the LMNA codon 482 and non-codon 482 mutation patient groups each had a high proportion of subjects with insulin resistance and hyperglycemia, diabetes, hypertension, menstrual irregularities, and elevated plasma triglycerides. Second, they found several key differences between the groups: non-codon 482 mutation carriers had milder lipoatrophy—three did not even have clinically apparent lipoatrophy—with higher BMI and larger waist circumference, and relative sparing of limb, trunk, and total fat stores. The milder lipoatrophy was confirmed by medical imaging studies, which showed that non-codon 482 mutation carriers had less atrophy of sc fat on the thighs and sparing of abdominal sc fat together with smaller visceral fat stores compared with codon 482 mutation carriers. The milder adipose redistribution in non-codon 482 mutation carriers was associated with older age at diagnosis, delayed onset of metabolic disturbances, and less severely depressed plasma leptin and HDL cholesterol. Also, sc lipomas and more severe acanthosis nigricans appeared to be more prevalent among codon 482 mutation carriers than among non-codon 482 mutation carriers.
The authors named the clinical syndrome in non-codon 482 mutation carriers "metabolic laminopathy" (MLP), a phenotype that is clearly distinct from typical FPLD2. In fact, MLP appears to more closely resemble FPLD3 (3), especially with respect to the pattern of adipose redistribution, which is less severe than in FPLD2 but is still associated with insulin resistance and a wide range of quantitative metabolic disturbances.
Of interest was whether subjects with these laminopathies met criteria for MetS. Some FPLD2 patients studied by Decaudain et al. (10) appeared not to have MetS, at least not by waist circumference criteria proposed by the International Diabetes Federation (IDF). In contrast, because of sparing of abdominal sc fat, all MLP (like most FPLD3) patients met the IDF waist circumference criteria for MetS. Parenthetically, FPLD2 patients provide an example of how increased waist circumference is not always crucial—perhaps thus not mandatory—to diagnose MetS, because FPLD2 patients met other defining criteria for MetS, including dyslipidemia, dysglycemia, and blood pressure, not to mention insulin resistance and diabetes. FPLD2 illustrates how criteria that depend on a mandatory waist circumference threshold might sometimes misclassify individuals as "unaffected" despite expanded visceral fat stores, which a simple tape measurement would not ascertain because of concomitant atrophy of abdominal sc fat.
Patients with partial lipodystrophy represent the ultimate in vivo"experiment-of-nature" with regard to human adipose tissue development and dysfunction. The differences in adipose tissue distribution between FPLD2 due to LMNA codon 482 mutations and MLP due to non-codon 482 mutations might lead to the identification of molecular pathways and programs that underlie varying degrees of lipodystrophy correlated with varying degrees of clinical involvement. An unresolved question is whether the metabolic disturbances in partial lipodystrophy develop secondarily to adipose tissue repartitioning or result from a direct effect of the mutant gene product. The phenomic studies of Decaudain et al. (10) suggest that the less extreme fat redistribution in MLP is associated with less extreme biochemical perturbations, although the temporal relationship between the onset of fat repartitioning and metabolic changes is not clear. However, the examples of FPLD2 and MLP indicate how combining genomic and phenomic perspectives can guide future experiments and perhaps improve understanding of common clinical entities, such as MetS or HIV-associated partial lipodystrophy. They also indicate that experiments in which the entire murine Lmna locus is ablated, even in a tissue-specific manner, might fail to capture the pathogenetic subtlety that is suggested by the clearly divergent phenotypes that result from single nucleotide mutations in human LMNA.
Thus, the concept of phenomics reinforces the value of clinical acumen and evaluative skills in phenotypic assessment. The current environment in molecular genetic research places a sizeable premium on the recapitulation of human phenotypes in animal models. Although these models have often proved helpful in moving our understanding of molecular pathways and targets forward, a large proportion of transgenic or knockout models fails to produce relevant and informative phenotypes for investigation of human disease. It is likely that careful phenomic evaluation of phenotypes in humans—as well as mice—will become an even more important component of biomedical research in the postgenomic world.
Footnotes
The author is supported by the Jacob J. Wolfe Distinguished Medical Research Chair, the Edith Schulich Vinet Canada Research Chair (Tier I) in Human Genetics, a Career Investigator award from the Heart and Stroke Foundation of Ontario, and operating grants from the Canadian Institutes for Health Research, the Heart and Stroke Foundation of Ontario, the Ontario Research Fund, and Genome Canada through the Ontario Genomics Institute.
Disclosure Statement: The author has nothing to disclose.
Abbreviations: FPLD2, Familial partial lipodystrophy subtype 2; MetS, metabolic syndrome; MLP, metabolic laminopathy.
Received September 17, 2007.
Accepted September 27, 2007.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
