Nucleoplasmic calcium regulates cell proliferation through legumain

Abstract

Background & aims: Nucleoplasmic Ca(2+) regulates cell growth in the liver, but the proteins through which this occurs are unknown.

Methods: We used Rapid Subtraction Hybridization (RaSH) to subtract genes in SKHep1 liver cells expressing the Ca(2+) buffer protein parvalbumin (PV) targeted to the nucleus, from genes in cells expressing a mutated form of nuclear-targeted PV which has one of two Ca(2+)-binding sites inactivated. The subtraction permitted the selection of genes whose expression was affected by a small alteration in nuclear Ca(2+) concentration.

Results: The asparaginyl endopeptidase legumain (LGMN) was identified in this screening. When Ca(2+) was buffered in the nucleus of SKHep1 cells, LGMN mRNA was decreased by 97%, in part by a transcriptional mechanism, and decreased expression at the protein level was observed by immunoblot and immunofluorescence. Treatment with hepatocyte growth factor increased LGMN expression. Knockdown of LGMN by siRNA decreased proliferation of SKHep1 cells by ∼50% as measured both by BrdU uptake and mitotic index, although an inhibitor of LGMN activity did not affect BrdU incorporation. A significant reduction in the fraction of cells in G2/M phase was seen as well. This was associated with increases in the expression of cyclins A and E. Furthermore, LGMN expression was increased in hepatocellular carcinoma cells relative to normal hepatocytes in the same specimens.

Conclusions: These findings suggest a new role for LGMN and provide evidence that nuclear Ca(2+) signals regulate cell proliferation in part through the modulation of LGMN expression. Increased expression of LGMN may be involved in liver carcinogenesis.

Figure 1. Schematic outline of the RaSH protocol

Tester and driver libraries were constructed, followed by digestion of only the tester library with XhoI. After hybridization, differentially expressed sequences were cloned into XhoI-digested vectors, resulting in a subtracted cDNA library enriched in genes displaying differential expression. By using the PV-NLS library as the tester and the PV-NLS-CD library as the driver, RaSH was used to produce a subtracted cDNA library enriched in genes up-regulated when nuclear Ca²⁺ is buffered (adapted from [17]). Down-regulated genes can be isolated using the PV-NLS-CD library as the tester and the PV-NLS library as the driver.

Figure 2. LGMN expression is decreased by buffering nuclear Ca²⁺

(A) Real Time quantitative PCR was used to measure the relative expression of LGMN mRNA in the SKHep1 liver cell line. LGMN mRNA was decreased by 97±2% in cells transfected with PV-NLS, relative to nontransfected controls (p< 0.0001). β-actin gene was used to normalize expression in both groups. The data are expressed as mean ± SEM of triplicate measurements and are representative of three separate experiments (p<0.05). (B) Immunoblot of whole-cell protein from SKHep1 cells 48 hr after infection demonstrates that LGMN protein expression is decreased after buffering nuclear Ca²⁺. Densitometric analysis confirms reduction of LGMN protein expression to 52±9% of controls (p<0.01). Expression of β-actin was used as a loading control. (C) Confocal immunofluorescence confirms decreased LGMN expression (green) in SKHep1 cells transfected with PV-NLS (red). Nuclei are identified by TO-PRO-3 staining (blue). Detection of the DsRed tag on PV-NLS confirms that it is localized to the nucleus. Results are representative of four independent experiments. Scale bar = 10µm. (D) Quantification of LGMN immunofluorescence shows that buffering nuclear Ca²⁺ reduces LGMN expression by 45% (11.1±0.4 in DsRed group versus 6.1±0.6 a.u. in PV-NLS-DsRed cells; p<0.001). (E) Luciferase assay shows that LGMN promoter activity is inhibited when nuclear Ca²⁺ is buffered (20.6±5.3 pLGMNprom versus 6.9±0.9 pLGMNprom + PV-NLS; p< 0.05). Data are expressed as mean fold increase ± S.E.M over empty vector in 5 separate experiments.

Figure 3. HGF stimulation increases LGMN expression

(A) Western blot analysis of total cell lysates prepared from control (non-stimulated) cells and cells stimulated with HGF (100 ng/mL) for the indicated time periods demonstrates a time-dependent increase in LGMN expression. (B) Bar graph shows the densitometric quantification of 4 separate experiments (p<0.05, one-way ANOVA). (C) LGMN localization is altered after HGF treatment. SKHep1 cells were stimulated as above and examined by confocal immunofluorescence. LGMN (red) is present in punctate structures that do not co-localize with the lysosomal marker Lamp-1 (green), in either control or HGF-stimulated cells. Nuclei are labeled by TO-PRO-3 (blue). Upon HGF stimulation, LGMN accumulates near the nucleus (arrows).

Figure 4. Knock down of LGMN inhibits cell proliferation

(A) Silencing of LGMN. SKHep1 cells were transfected with 20 nM siRNA for LGMN, and incubated for 72 hrs. Immunoblots demonstrate that LGMN but not scrambled siRNA knocks down LGMN expression. α-tubulin serves as a loading control. Densitometry confirms reduction of LGMN to 21.5±3.2% of nontransfected controls (p<0.0001). Results are representative of three separate experiments. (B) Knockdown of LGMN decreases BrdU incorporation in SKHep1 cells to 50.7±8.1% of controls (p<0.0034). (C) Mitotic SKHep1 cells were identified by confocal imaging of phospho histone-3 labeling measured 72 hrs after knockdown of LGMN. The mitotic index is decreased to 2.2±0.7% of control in cells in which LGMN is silenced (p<0.0001). A total of 300 cells were examined in three separate experiments. (D) Inhibition of LGMN activity with MV026630 at either 25 or 50 µM does not alter BrdU uptake in SKHep cells (p>0.05).

Figure 5. Cell cycle kinetics after knockdown of LGMN

(A) Representative FACS cell cycle profiles of non-transfected (NT) and siRNA-transfected SkHep cells 72 hr after treatment with scrambled or LGMN siRNA. (B) In cells in which LGMN was silenced, there was a reduction in the fraction of cells in G₂/M phase (7.4±10.9% in LGMN siRNA versus 25.6±3.5% in non-transfected; mean+SD; p<0.05), without a significant increase in the fraction of cells in G₁ or S phase. Cell cycle profiles were not changed in cells transfected with scrambled siRNA. Data are mean of 3 independent experiments. (C) Knockdown of LGMN does not induce apoptosis, as measured by caspase-3 activity. Staurosporine (500 nM) was used to induce apoptosis as a positive control for caspase-3, and a caspase-3 inhibitor was used as a negative control. Bar graph shows that caspase-3 activity was not increased in response to knockdown of LGMN (p>0.05, by one-way ANOVA). Results are representative of four independent experiments (*p<0.001).

Figure 6. Progression of cell cycle after knockdown of LGMN

(A) Immunoblot of total protein from SKHep1 cells tests expression of various cell cycle regulatory proteins 72 hr after knockdown of LGMN. (B) Densitometric analysis summarizes the results of the western blots. There was a significant increase in expression of cyclin A (*p<0.01) and cyclin E (**p<0.05) in cells transfected with LGMN siRNA relative to non-transfected. Expression of α-tubulin was used as an internal control for protein loading. Data are mean of three independent experiments.

Figure 7. Expression of LGMN is increased in hepatocellular carcinoma (HCC)

Confocal immunofluorescence images were obtained from paraffin-embedded surgical specimens of tumors from patients with HCC. Immunohistochemical staining was performed to determine expression of LGMN (green) in tumor cells and in nearby normal hepatocytes. To-Pro-3 was used to identify cell nuclei (blue). (A) Low-power (10×) image of carcinoma cells and normal hepatocytes in the same field of view shows that LGMN staining is increased in the HCC. (B) Higher magnification (63×) images confirm increased expression of LGMN and show that it is distributed throughout the cytoplasm in HCC (Scale bar = 30 µm). Findings are representative of what was observed in 3 fields each of specimens from 5 separate patients. (C) Quantification of the average fluorescence in normal and HCC affected areas in the same specimen shows a significant increase in LGMN expression in the carcinoma cells (53.9±4.8 a.u.) as compared to normal hepatocytes (38.3±3.1 a.u.; p<0.01, paired t-test).