Inhibition of protein farnesyltransferase activity by FTIs has been shown to improve abnormalities in nuclear morphology in cells from subjects with HGPS and from animal models of the disease 
. FTIs also improve the progeria-like phenotypes of mouse models of HGPS 
. Presumably, blocking farnesylation of progerin, the truncated prelamin A expressed as a result of the G608G LMNA
mutation which is responsible for most cases of HGPS, renders it less toxic. However, the molecular mechanisms by which progerin exerts its toxicity and whether FTI treatment reverses specific molecular defects induced by progerin remain largely unknown. We have now identified a defective lamin A-Rb signaling network in cells from subjects with HGPS and demonstrated that treatment of cells with an FTI reverses abnormalities in the expression of genes encoding proteins in this network.
We compared our gene expression profiles to those of previous studies that have examined gene expression in HGPS and found very little overlap (3.5% overlap of differentially expressed genes with Scaffidi and Mitseli 
and less than 1% with Csoka et al. 
). Csoka et al.
used fibroblasts from only three subjects ages 8 to 14 years of age whereas we used cultured fibroblasts from five subjects with HGPS obtained at earlier ages (age 1 to 4 years). As the growth rate and proliferation potency of fibroblasts from subjects with HGPS decrease with cellular age in vitro
and the donor's age, this could partially explain the differences between our results. Importantly, the average age of death in HGPS is 12–15 years, which suggests that the cellular phenotype may be severely altered in older children 
. Scaffidi and Misteli 
generated immortalized normal fibroblasts that overexpressed progerin and examined gene expression in only two lines, making a comparison difficult given the very different system and the low statistical power from using only two biological replicates. Separately, Fong et al. 
used RT-PCR to examine expression of several genes selected from these previous studies in normal fibroblasts transfected with antisense oligonucleotides that increase LMNA
alternative RNA splicing to generate progerin. We found that four of seven genes they tested were similarly differentially expressed in the fibroblasts we used from subjects with HGPS (Table S2
). An earlier study by Ly et al
used fibroblasts from three donors with HGPS of age 8 to 9 years and found 61 genes differently expressed in young subjects compared to old subjects and with HGPS subjects. Of these 61 genes, 49 were differentially expressed in HGPS versus young individuals 
. Comparison of these 49 genes against the 352 genes differentially expressed in HGPS identified in our study indicated an overlap of 22 genes. An additional study from Park et al 
compared gene expression in fibroblasts from only one subject with HGPS to that of cultured fibroblasts in replicative senescence; no statistical criteria were applied. Although HGPS is an extremely rare disease, to the best of our knowledge we have used the most biological replicates so far in any study of genome-wide gene expression in actual patient material. We also validated our microarray data using information from the literature, by screening databases and by performing RT-PCR for selected genes. We confirmed by RT-PCR that Rb mRNA levels were decreased in HGPS cells in comparison the control cells. No change in Rb gene expression level was reported in previous microarray analyses on HGPS cells or HGPS cell models 
Most importantly, rather than providing a descriptive list of genes with altered expression levels, our methods identified a novel signaling network starting with the lamin A-Rb interaction that is altered in HGPS.
The potential link between progerin and Rb in the orchestration of cell cycle and proliferation defects was previously suggested by studies demonstrating that lamin A and C interact with Rb 
. Rb is a tumor suppressor and major cell cycle regulator that, in its hypophosphorylated state, binds to and inhibits the E2 factor (E2F) family of transcription factors required for cell cycle progression 
. Upon hyperphosphorylation of Rb by cyclin/cyclin-dependent kinase complexes, E2F is released to initiate S phase. A role for lamins in this process is further suggested by the finding that hypophosphorylated Rb is tightly associated with lamin A/C-enriched nucleoskeletal preparations of early G1 cells 
. In HGPS cells there is evidence for a significant reduction in hyperphosphorylated Rb in HGPS fibroblasts (Dechat et al. 2007). In our experiments, we found a decreased level of Rb expression at the mRNA and protein levels in HGPS (see and ).
Previously, a small fraction of lamin A/C was shown to colocalize with Rb and E2F1 in nuclear foci during G1-phase 
. In HGPS cells, we could not detect lamin A foci. Whether Rb foci remain associated with E2F1 remains to be further investigated. In HGPS cells, our microarray analysis indicated a decrease in Rb levels without changes in E2F1 levels. We further investigated whether E2F1 levels might be altered by indirect immunofluorescence microscope and observed no changes in the nuclear intensity of the E2F1 signal between HGPS and normal fibroblast cells (data not shown). Previously, the localization of Rb and E2F1 in cells from Lmna−/−
mice indicated that while Rb levels were reduced, a portion of Rb remained localized to nuclear foci but displayed reduced overlap with the E2F1 signal 
. The expression levels of E2F1 in Lmna−/−
cells in comparison to wild type cells has not yet been reported. However, the lack of association between Rb and E2F1 in nuclear foci in the absence of lamin A/C could indicate an alteration in Rb control of E2F1 activity 
. Further studies will be needed to understand how expression levels of lamin A/C, Rb, and E2F1 might affect each other's expression levels, interactions and subnuclear localization. In addition to its role in cell cycle control, Rb also regulates cellular differentiation. In the absence of lamin A/C in muscle cells derived from Lmna−/−
mice, reduced levels of Rb and other transcription factors regulating muscle cell differentiation was reported 
. This study provide an additional link between lamin A/C-Rb and its potential role in muscle cell differentiation 
Immunohistochemical analysis of Rb distribution did not reveal any obvious alteration besides a decreased Rb signal in some nuclei of cells from subjects with HGPS (Figure S7
). Rb was also detected in the most dysmorphic nuclei exhibiting a strong progerin staining, indicating the accumulation of high levels of progerin protein (Figure S7
) as reported previously 
. Based on these previous studies, decreased Rb expression or phosphorylation status in HGPS cells appears to be implicated in the deregulation of proliferation. Whether the interaction of A-type lamin with Rb is impaired in HGPS cells remains to be further addressed. However, our present study together with previously published observations support the potential implication of a defective lamin A/C-Rb signaling network in the HGPS cellular phenotype. In support of our finding, a similar potential link between an altered Lamin A/C-Rb signaling was previously reported in cells derived from Lmna−/−
mice model 
. In cells lacking lamin A/C the levels of Rb was decreased while its mRNA level remained unchanged in comparison to wild type cells 
. These findings suggested that the Rb was unstable in the absence of lamin A/C. In HGPS cells, we observed a decrease in both mRNA and protein levels of Rb. Our findings indicate that the lamin A/C-Rb signaling might be altered differently depending on the status of A-type lamins in cells (absence of A-type lamins expression versus expression of an abnormal lamin A variant, progerin). Further studies are required to understand the molecular mechanisms driven by alterations in A-type lamins on Rb signaling pathways.
FTI treatment of cells from subjects with HGPS leads to a significant reversal of the abnormal expression of genes encoding proteins in the lamin A-Rb signaling network. This is consistent with the improved phenotypes that result from blocking protein farnesylation in cellular and animal models of the disease 
. Defective Rb activity appeared to be responsible for the repression of the large set of downstream transcription factors and regulators. To evaluate the impact of FTIs, we also investigated FTI-induced gene expression changes in fibroblasts from unaffected controls. Remarkably, Rb appeared again in this dataset as a key regulator of the downstream events. Not surprisingly, FTI-treated cells from subjects with HGPS demonstrated a near total reversal (99%) of abnormally expressed genes in comparison to FTI-treated control cells. Our finding suggests that a potential correction of the HGPS phenotype does not necessarily require a full inhibition of protein farnesyltransferase but rather points to the importance of normalizing Rb expression levels and function.
The results of our study further support the potential use of FTIs as therapeutic agents in HGPS. A phase II clinical trial treating children with HGPS with lonafarnib, the FTI we used in this study, was initiated in spring of 2007 
. Because of known toxicity, the dose of lonafarnib administrated to these children may generate tissue concentrations below those necessary to obtain the effects we observed in vitro
. Therefore, in human subjects, other drugs in addition to an FTI may be necessary to inhibit protein prenylation to a similar extent as in cultured cells. One possibility is to add a statin and an aminobisphosphonate, which have been shown to inhibit prelamin A prenylation and improve the phenotype of mice deficient in Zmpste24 
. Furthermore, our findings suggest that screening candidate molecules against targets we identified in the lamin A-Rb signaling network could identify novel therapeutic approaches to treat HGPS. Proteins in this network could also serve as biomarkers for HGPS.
Because the lamin A-Rb interaction is a new signaling axis in the pathogenesis of HGPS, we investigated its potential role in physiological aging. We found that LMNA
expression appeared to be inversely modulated in the elderly. These observations were similar to our findings in cells from subjects with HGPS. Although lamin A and lamin C levels were not altered, an abnormal prelamin A variant, progerin, is expressed and Rb expression is decreased. In addition to changes in the expression of A-type lamins, several studies suggest that abnormal prelamin A may accumulate in normal aging 
. Our results, therefore, suggest that therapeutic approaches to re-establish a proper lamin A-Rb signaling network may be beneficial in preventing complications of physiological aging.