Abstract
Background/Aim: Nucleic acid metabolism is biochemically compartmentalized to the nucleus. Thus, it is necessary to define the proteome of the various macromolecular structures within this organelle. Materials and Methods: We isolated the nuclear matrix (NM) fraction from rat liver by sequential centrifugation steps at 13,000 rpm, staggered between endogenous nuclease treatment for 2 h at 37°C, followed by high-salt (H.S.; 2.0 M NaCl) and non-ionic detergent extractions (0.1%- or 1.0% Triton X-100) to eliminate the bulk of chromosomal DNA/RNA, histone proteins and the nuclear envelope (NE). Results: Integrity of the NM and NE structures was confirmed by electron microscopy. Next, we analyzed the NM proteome on a 20% polyacrylamide gel using the PhastSystem. We observed the absence of histone proteins and the characteristic presence of the lamins by Coomassie blue staining. By contrast, upon silver staining, following electrophoretic separation with a Tris-Borate-EDTA buffer, we observed the NM-associated nucleic RNA and protein-free ADP-ribose polymers. While polymers are found in much lower concentration than RNA in NM, they were purified by affinity chromatography on boronate resin prior to electrophoresis. We observed the electrophoretic resolution of free ADP-ribose chains (5-25 units) by silver staining. Conclusion: The significance of our observations to cancer studies and carcinogenesis is discussed.
- Nuclear matrix
- proteomics
- electrophoresis
- PhastSystem
- lamins
- histones
- poly(ADP-ribose)
- boronate resin
- affinity chromatography
- epigenetics
- nuclear envelope
- nucleoli
- chromatin
NM, nuclear matrix; NE, nuclear envelope; NPC, nuclear pore complex, L.S., low salt; H.S., high salt; rpm, revolutions per minute.
The nuclear matrix (NM) is a non-aqueous macromolecular complex of mostly RNA and protein that spans the nucleoplasmic space of eukaryotic cells (1) and provides a rigid scaffold that contributes to the quasi-spherical shape of this prominent cell organelle (2). Even though the nuclear matrix (NM) has not been fully confirmed as a functional biological entity in vivo (3), it has been the subject of intense investigation and has been defined as a fixed platform or “scaffold” for the topographical organization of pivotal metabolic chromatin processes such as DNA replication (1, 4, 5), gene expression (1, 6, 7) and DNA repair (1, 8, 9), amongst others. Interestingly, after 30 years of some controversy regarding the actual biological significance of the NM unit in vivo, recent microscopic visualization of the nucleoplasmic scaffold has persuaded the scientific community to accept the actual existence of the NM as the architectural skeleton of the cell nucleus (10). Not surprisingly, it has also recently been identified as a rich source of potentially important cancer protein biomarkers (11) in human neoplasia.
Ever since the first report of NM isolation from rat liver cells by Berezney and Coffey (12), it was clear that this macromolecular structure was made off 98.4% protein, 1.2% RNA, 0.5% phospholipids and 0.1% DNA. In fact, the main polypeptide components of the NM were identified as three abundant acidic proteins with a molecular mass of 60-70 kDa (12). Although this sub-nuclear structure seemed to be always devoid of the bulk of nucleic acids (DNA and RNA), as early as 1979, it was suggested that (13) the NM might be crucial for the control of gene expression.
Regarding the molecular composition of the nuclear scaffold, it has been shown that at least 400 polypeptides are associated with the NM (14). While more than 50% of these proteins may be temporarily associated with this sub-nuclear unit, depending on the stage of the cell cycle, it is now clear that the NM is a significantly dynamic structure that facilitates the proper compartmentalization of the replisome (1, 4, 5), the transcriptome (1, 6, 7) and the repairosome (1, 8, 9).
More importantly, the most prominent molecular components of the NM, namely lamins A, B and/or C, have recently been implicated with specific human diseases, grouped together in a series of pathological anomalies identified as laminopathies (15). This timely report, recently published (15), carefully classified the various mutations that occur in genes encoding the intermediate filament nuclear lamins, as well as associated polypeptides, as they correlate with various pathologies such as cancer, progeria, aging, autosomal dominant Emery-Dreifuss muscular dystrophy, Charcot-Marie-Tooth disease, autosomal dominant leukodystrophy and partial lipodystrophy, etc. (15).
Therefore, new rapid and reproducible protocols for the consistent isolation of the nuclear matrix from mammalian tissues are required. Due to the fact that the quantity and identity of proteins co-purifying with the NM fraction varies from cell type to cell type and with the stage of the cell cycle (15, 16), new NM isolation procedures must include a key methodological step that confirms the presence of lamins A, B and C as the markers of this structure, concomitant with the virtual absence of histones, the most abundant proteins typically found in eukaryotic chromatin (7, 17) and which comprise up to 50% of the proteome of the cell nucleus.
Herein, we report the reproducible isolation of the NM fraction from rat liver nuclei, followed by the electrophoretic analysis of the proteome and nucleic acids co-purifying with the NM macromolecular complex, on separate 20% SDS gels using the PhastSystem. We consistently observed a Protein/RNA/poly(ADP-ribose) fraction that predominantly contains the intermediate filament lamins (60-70 kDa), completely free of histone proteins.
Materials and Methods
Materials. Histones and β-NAD+ were purchased from Sigma Chemical Company (St. Louis, MO, USA). Bio-Rex 70, molecular weight markers and 10 X SDS-Tris-Glycine and 10 X Tris-Borate-EDTA stock solutions were obtained from Bio-Rad (Richmond, CA, USA), Pre-casted 20% homogeneous polyacrylamide gels, agarose buffer strips and PhastSystem sample applicators were from GE Healthcare (Piscataway, NJ, USA). All other reagents were reagent grade.
Isolation of nuclei. Rat liver nuclei were obtained from Sprague-Dawley Rats (Charles Rivers Lab, San Diego, CA, USA) and immediately homogenized in a 250 mM sucrose solution containing TKM buffer (50 mM Tris-HCl, pH 7.5 at 4 °C; 25 mM KCl and 5 mM MgCl2). The homogenate was filtered through multiple layers of cheesecloth. The mixture collected was further diluted 1:3 (v/v) with 2.3 M sucrose in TKM buffer. Next, the suspension was centrifuged at 39,000 rpm at 0-4°C. The liver nuclei contained in the pellet were further used as a source for sub-nuclear fractions.
Fractionation of nuclei. Rat liver nuclei were fractionated as previously described (18). Briefly, nuclei were incubated at 37°C for 2 h to digest the DNA with endogenous nucleases. Nuclei were subsequently centrifuged at 780 x g for 15 min at 4°C. The pellet was diluted in 10 mM Tris-HCl, pH 7.4 (Buffer A) containing 200 μM MgCl2. Three extractions were carried out with buffer A to remove chromatin DNA. The bulk amount of chromatin remaining, e.g., histone proteins, was extracted with buffer A containing 2.0 M NaCl. Next, the NM was obtained by the treatment of the remaining material with 0.1 % Triton X-100, 5 mM MgCl2. After washing the pellet twice in buffer A, containing 5 mM MgCl2 to remove the residual Triton X-100, nuclear matrices were resuspended in buffer A containing 5 mM MgCl2. The different fractions collected were stored at -20°C until used. The NM fraction was applied to a 20% polyacrylamide gel and electrophoresis was carried out as described below, on the PhastSystem.
Preparation of sub-nuclear samples for electron microscopy. A standard protocol for electron microscopy was followed. First, the samples were fixed overnight with a solution constituted of 2.5% glutaraldehyde in 125 mM PIPES buffer (piperazine-N,N’-bis(2-ethanesulfonic acid)), pH 7.3. In addition, purified nuclei and nuclear matrices also contained buffer A (vide supra) with 200 μM MgCl2. Subsequently, samples were post-fixed in 1% OsO4. Dehydration of the samples was accomplished by a series of treatments with graded ethanol solutions. The fixed material was infiltrated with different ratio mixtures of propylene oxide/epon 812. The samples were then sectioned via Leica ultracuts with a diamond knife. Finally, ten and three hundred nanometer sections were stained with orange acetate and Sato’s lead. The different sub-nuclear fractions were viewed with a Zeiss EM 910 electron microscope (Zeiss Instruments, Thornwood, NY).
Poly (ADP-ribose) purification. ADP-ribose polymers were purified by affinity chromatography on dihydroxyboronyl-Bio Rex 70 (DHB-B) as described elsewhere (19, 20). Briefly, nuclear matrices containing the NM proteome and nucleic acids were dissolved in 250 mM ammonium formate, pH 9.0 containing 6.0 M guanidine-HCl (10 ml). The samples were loaded onto a 1.0 ml DHB-B column previously equilibrated with the affinity chromatography loading buffer at 4°C. Then the column was washed three times with 10 ml of ice-cold 250 mM ammonium formate, pH. 9.0 and the protein-free ADP-ribose polymers were eluted with water at 37°C. The eluate was lyophilized and used for 20% polyacrylamide gel electrophoresis on the PhastSystem.
Electrophoresis and staining of the NM proteome, nucleic acids and protein-free ADP-ribose polymers. The protein composition of each sub-cellular fraction was first analyzed by SDS-polyacrylamide gel electrophoresis as described by Laemmli (21)Reference) with 0.1% SDS, 12% polyacrylamide (Figure 2) and 4 M Urea. For all other experiments, the nuclear matrix proteins were electrophoresed under SDS-denaturing conditions using pre-casted 20% polyacrylamide gels (GE Healthcare (Piscataway, NJ, USA)) and 50 mM Tris-Glycine buffer, pH 7.8 with the PhastSystem. Separation parameters were as follows: 250 V, 10 mA, 3.0 W, 15°C, for 90 minutes. Sample loading buffer consisted of 50 mM Tris-Glycine buffer, pH 7.8 from Bio-Rad (Richmond, CA, USA) containing bromophenol blue as a marker and high molecular weight standards, also from Bio-Rad. The gels were fixed in methanol/acetic acid and stained with Coomassie blue for protein. Nucleic acids were also electrophoresed in homogenous 20% polyacrylamide gels and the electrophoresis buffer used was 89 mM Tris-Borate pH 8.3 containing 2.0 mM EDTA. Silver staining for RNA and poly(ADP-ribose) and high specificity was performed according to the commercial specification of the Protostain system from National Diagnostics (Atlanta, GA, USA).
Results and Discussion
Recently, significant interest has been pointed towards the NM proteome as a potential source of cancer biomarkers (10, 16). The rationale is that the mitotic dependence of the NM integrity may be directly involved in the regulation of key cell division metabolic events (15, 16) such as DNA replication (1, 4, 5) or cell cycle-dependent gene expression (1, 6, 7) patterns. At the same time, based on the cell type used as a source for NM purification (13), as well as the actual cell cycle status (220) and, more importantly, the actual protocol of isolation (15, 16), the proteome generated may have varying polypeptide compositions.
For example, while the use of (NH4)2SO4 for the high ionic strength extraction of histone proteins may yield a hepatocyte NM proteome with over 500 proteins (16), other methods may result on an average of 400 polypeptides (14). Nonetheless, it must be re-emphasized that, regardless of the NM isolation protocol used, all reports coincide with the conclusion that the most abundant polypeptides in this fraction are lamins A, B or C (1, 13, 14, 16), a class of proteins (23) that belongs to the family of intermediate filament V proteins (23-25). In addition, it must be noted that mutations in the primary sequence of these proteins are currently associated with NM dysfunctions (13) that may lead to several diseases defined as human laminopathies (15). Therefore, it is clear that all methods of NM isolation must corroborate the presence of the lamins (60-70 kDa) as markers of purity, while the extracts must be devoid of histones (vide infra).
Figure 1 displays the morphological integrity of the NM fraction that we isolated from rat liver as described in the Materials and Methods section. Panel A shows the electron microscope structure of intact rat liver nuclei. Panel B shows the morphological appearance of the nuclear product after endogenous nuclease digestion to remove the bulk of DNA. Panel C displays the NM structure still containing the nuclear envelope (NE), after the 2.0 M NaCl extraction of the histone proteins. By contrast, the NM proteome obtained (with small amounts of RNA), after the elimination of the double lipid bilayer with 1.0% Triton X-100, is shown on Panel D. Our morphological results under the electron microscope are identical to NM’s isolated from other tissues (11), as well as single cell organisms, such as Physarum polycephalum (22).
Figure 2 shows the protein composition of all rat liver nuclei sub-components that were collected during the sub-nuclear fractionation as described above, including the low-salt (L.S.), high-salt (H.S.) and wash solutions. On Figure 2, panel A, lanes1 through 6 display the molecular weight markers (M), isolated nuclei, the endonuclease digestion products and nuclease digested nuclei, low-salt chromatin extract and the Nuclear Matrix I fraction, still containing the NE and all histone proteins, respectively. By contrast, Figure 2-panel B shows in Lanes 1 and 2 the molecular weight markers and an aliquot of intact rat liver nuclei again. Lanes 3 through 7 display the 2.0 M (H.S.) extract, the Nuclear Matrix II (without histone proteins) with NE and Nucleoli, the NE (Triton X-100 Material), the proteins in the washing solution and the purified Nuclear Matrix III (Lamins) that was further processed for the rapid and reproducible 20% polyacrylamide gel analysis using the PhastSystem shown on Figures 3 and 4 below. Interestingly, as shown on panel A, lane 3, most DNA-binding proteins were extracted after the endogenous nuclease digestion of rat liver nuclei. By contrast, the histone proteins were preferentially extracted via the L.S. (lane 5, panel A) and H.S. (lane 3, panel B) NM purification steps. Furthermore, the multi-subunit NE complex, as well as integral NE proteins, were extracted with the Triton X-100 step (lane 5, panel B). Thus, as expected, the nuclear matrix III (lane 7, panel B) was almost exclusively constituted of the protein lamins. However, the process of methodological confirmation that the final nuclear matrix proteome was predominantly constituted of histone proteins and completely devoid of histones was both tedious and time consuming. Therefore, we proceeded to speed-up the process of methodological confirmation of these two key molecular parameters using pre-casted and commercially available 2×2 inch 20% polyacrylamide gels and the PhastSystem followed by Coomassie blue staining in less than two hours as shown below.
Figure 3 shows that the main Coomassie blue stained the polypeptides observed, run in the range of 60-70 kDa molecular weight markers, with a 20% polyacrylamide gel under SDS denaturing conditions, which is consistent with the significant abundance of the lamins as markers of NM purity. Lanes 2 and 3 show the lamins just barely migrating 5 mm below the origin due to the high concentration of polyacrylamide (20%). While a few other lower molecular weight proteins, typically associated with the NM fraction, are also present in our samples; their relative abundance is not high enough for Coomassie blue staining. Instead, we clearly observed that all polypeptides with a molecular weight higher than 80 kDa do not migrate into the gel and remain as a single protein band at the origin of the stacking gel. To further confirm the absence of histones, Figure 3-lane 4 shows the expected electrophoretic migration of a mixture of histone H1 and core histones, which are not detectable on lanes 2 and 3. Therefore, in order to further confirm the quantitative removal of histone proteins from the NM by the 2.0 M NaCl extraction (vide supra), we run another 20% polyacrylamide gel in parallel, under SDS denaturing conditions as well, but we stained with silver nirate as described in the Materials and Methods section. We observed a smear of silver stained RNA molecules that span the lane completely from the origin of the gel to the bromophenol blue marker at the bottom of the gel (Figure 4, lanes 2 and 3). Lane 4 shows the histone molecular weight markers for comparison.
The presence of a significant amount of silver-stained material is consistent with the presence of short RNA molecules known to be associated with the NM (1-3, 11-13) and perhaps other less abundant nucleic acids, such as the covalently bound ADP-ribose polymers (18, 264) found in the NM. This explanation would also be consistent with the fact that ADP-ribose chains are chemically released from the protein acceptors by the mild alkaline conditions (27-30) of electrophoresis. Due the importance of the NM in telomere biology (25) and DNA replication (18) and repair (25), we proceeded to seek whether our system would also be useful to determine the size distribution of ADP-ribose polymers (31) in the NM fraction.
Figure 5 shows the silver-stained profile of the ADP-ribose chains co-purifying with the NM proteome after affinity chromatographic purification on DHB-Bio Rex (2019). The free ADP-ribose polymers obtained appear as a smear of silver stained oligonucleotides (Figure 5, lanes 1 and 2) in a region of the electrophoregram that corresponds to ADP-ribose chain lengths oscillating between 5 and 25 ADP-ribose units. Also shown on lanes 1 and 2 is an aliquot of 50 ng ADP-ribose pentamer that was added “to spike” the samples with an internal control for electrophoretic migration reference. Interestingly, using our protocol, we did not see a strong silver band at the origin of the gel, consistent with the notion that no highly branched polymers (31) are synthesized covalently bound to nuclear matrix proteins. Lanes 3 and 4 of Figure 5 show the electrophoretic migration of a pentamer and a monomer of ADP-ribose, respectively.
In conclusion, we have demonstrated that the utilization of homogenous 20% polyacrylamide gels and the PhastSystem can be a useful tool to quickly monitor for the presence of lamins and the absence of histone proteins to confirm the expected protein composition of a nuclear matrix fraction. In addition, we also show that an identical gel can be used to determine the level of NM associated protein-poly(ADP-ribosyl)ation. The presence of these unique polynucleotides, covalently bound to NM protein, becomes even more relevant in light of the recent report of Gonzalez-Suarez et al. (25) who demonstrated that the NM is definitely involved in the DNA-damage response pathway and also telomere biology, which eventually connects the NM scaffold as a relevant structure in the early stages of carcinogenesis as suggested by Albrethsen et al. (11), as well as Barboro et al. (16) (vide supra). Likewise the approach described in this report may also become very useful in detailed and careful characterization of the NM, NE and nuleoli proteomic samples acquired from individuals with cancer (11) containing covalently bound modifying groups that alter their overall chemical structure (32) and may epigenetically modify their function. Interestingly, while this study was in the final stages of preparation, a timely report in the Journal of Cell Biology (33) discussed the significance of proteome-bound ADP-ribose polymers in cell and nuclear architecture. In fact, this recent careful analysis confirms previous suggestions (27, 34-36) that the covalent and reversible protein-poly(ADP-ribosyl)ation (37, 38) flows from chromatin and nuclear structures, including the NM (39), to cytoplasmic organelles (35). Similarly, the metabolism in both germ cells (27) and somatic cells (34, 35), including cervical cancer He-La cells, shows re-distribution of ADP-ribose polymers in the nucleoli (36), upon genotoxic damage (40), along with the enzyme (36) responsible for their synthesis (PARP-1) [EC 2.4.2.30]. Interestingly, our observations and interpretations were reproduced 2012 (41) and further confirmed even more recently, in 2014 (33).
Clearly, our understanding over the interplay between the genomics and proteomics of the protein-poly(ADP-ribosyl)ation pathways in human cells shall prove instrumental in the diagnosis, prognosis and treatment of cancer.
Footnotes
-
↵* Present Address: Department of Biology, University of Texas at Arlington, Arlington, TX, 76019, U.S.A.
-
↵** Present Address: Internal Medicine Residence Program; Mayo Clinic; 4500 San Pablo Road South, Jacksonville, FL, 32224, U.S.A.
-
↵*** Present Address: Cardiovascular and Pulmonary Research Branch, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD, 20892, U.S.A.
- Received July 10, 2014.
- Revision received August 4, 2014.
- Accepted August 6, 2014.
- Copyright© 2014, International Institute of Anticancer Research (Dr. John G. Delinasios), All rights reserved