Abstract
Cyclin T2a was recently identified as one of the regulatory subunits of the cdk–cyclin complex P-TEFb, the most studied positive factor in the regulation of transcription elongation. By fluorescent in situ hybridization (FISH), the gene codifying for cyclin T2a has been mapped on human chromosome 2q21. This locus also has been linked to different forms of myopathy. By use of a new specific antiserum raised against cyclin T2a, the immunohistochemical pattern of expression of cyclin T2a in human tissues has been examined and compared to that of cyclin T1, described in the previous report. The observation that immunohistochemical expression of cyclin T2a was high in skeletal muscle cells, whereas it was undetectable in two cases of centronuclear myopathy, together with its chromosomal location, suggests an involvement of the cdk9–cyclin T2a complex in this disease.
Transcription elongation in eukaryotic genes is a complex process that involves a number of regulatory factors. A common rate-limiting step in the transcription of inducible genes is the release of RNA polymerase II (RNAPII) molecules from stalled elongation complexes that accumulate shortly after the initiation of RNA synthesis (Reines et al. 1996). This step is controlled by both negative and positive factors. Negative factors, such as factor 2 (Xie and Price 1996, 1997, 1998) and 5,6-dichloro-1-β-d-ribofuranosyl-benzimidazole (Hartzog et al. 1998; Wada et al. 1998), cause abortive elongations in which only short transcripts are generated. Positive factors enable RNAPII to overcome the promoter proximal pausing and premature termination and to enter productive elongation. The most studied positive factor is the transcription elongation factor b (P-TEFb) (Marshall and Price 1992, 1995). P-TEFb promotes the transition into productive elongation by phosphorylating the carboxyterminal domain (CTD) of the largest subunit of RANPII (Marshall et al. 1996). Drosophila P-TEFb has been cloned recently and identified as a cyclin/cyclindependent kinase (cdk) pair, able to phosphorylate CTD in vitro (for review see Yankulov and Bentley 1998). Most well-known cyclin–cdk pairs play a fundamental role in regulating the cell cycle (MacLachlan et al. 1995), but some have been found involved in other cellular events, as in the case of P-TEFb. Human P-TEFb has been identified as a cdk (cdk9) paired with a cyclin subunit (cyclin T) (Peng et al. 1998; Ping and Rana 1999). The cdk9 gene was cloned first and was identified as a cdk-like kinase with unknown function (PITALRE) (Graña et al. 1994). On the other hand, the cyclin T subunit can be provided by two related genes: cyclin T1 and T2. In addition, cyclin T2 has two forms, T2a and T2b, probably arising from alternative splicing of the primary transcript (Peng et al. 1998). Human cdk9 mRNA, as well as cyclin T1, T2a, and T2b mRNAs, is expressed ubiquitously, which is consistent with the general requirement of P-TEFb in transcription elongation. The presence of multiple cyclin subunits for a single cdk partner is a well-described phenomenon (MacLachlan et al. 1995). However, different roles for these three cyclins are possible in specific situations, as has been shown for cyclinT1–cdk9 complex upregulation during T-cell activation (Garriga et al. 1998).
Here we describe the chromosomal mapping of the human cyclin T2a gene and the characterization of a new polyclonal antiserum raised against its gene product. By use of this antiserum, the immunohistochemical pattern of expression of cyclin T2a in normal human tissues was examined and was compared with that of cyclin T1, described in the previous report (De Luca et al. 2001). The data on chromosomal mapping and the immunohistochemical expression in human adult muscle cells and in some cases of myopathy, together with earlier data on the cdk9-cyclin complex activity (Bagella et al. 1998), suggest a possible role for cyclin T2a in muscle differentiation.
Materials and Methods
Chromosomal Mapping
Metaphase spreads from phytohemagglutinin-stimulated lymphocytes of a healthy female donor were prepared as described (Bell et al. 1995). FISH and detection of immunofluorescence were performed essentially as previously described (Bell et al. 1995). A 2.2-kb cyclin T2a cDNA in pcDNA3 was biotinylated by nick-translation in a reaction containing 1 μg DNA, 20 mM each of dATP, dCTP, and dGTP (Perkin-Elmer; Norwalk, CT), 1 mM dTTP (Perkin-Elmer), 25 mM Tris-HCl, pH 7.5, 5 mM MgCl2 (Sigma; St Louis, MO), 10 mM β-mercaptoethanol (Sigma), 10 mM biotin-16-dUTP (Boehringer Mannheim; Mannheim, Germany), 2 U DNA polymerase 1/DNase 1 (GIBCO BRL; Gaithersburg, MD), and H2O to a total volume of 50 μl. The probe was denatured and hybridized to metaphase spreads overnight at 37C. Hybridization sites were detected with fluorescein-labeled avidin (Oncor; Gaithersburg, MD) and amplified by addition of anti-avidin antibody (Oncor) and a second layer of fluorescein-labeled avidin. Chromosomes were counter-stained with diamidino-2-phenylindole (DAPI) and observed with a Zeiss Axiophot epifluorescence microscope equipped with a cooled charge-coupled device camera (Photometrics; Tucson AZ) operated by a Macintosh computer work station. Digitized images of DAPI staining and fluorescein signals were captured and merged using Oncor Image version 1.6 software.
Normal Tissues and Cell Lines
Normal tissues from autopsy, as well as tissue from two patients with myopathy, were obtained from the Department of Surgical Pathology of the Second University of Naples, Italy. Tissues were formalin-fixed and paraffin-embedded. Representative sections of each specimen were stained with hematoxylin-eosin and examined by a pathologist to confirm the histological preservation of the microanatomic structure. For each tissue examined, at least three specimens from two different individuals were analyzed.
Human tumor cell lines SAOS-2 (osteosarcoma), T98G (malignant glioma), and HeLa (cervical carcinoma) were obtained from the American Type Culture Collection (Manassas, VA) and were maintained in culture in Dulbecco's modified Eagle's medium (DMEM) complemented with 10% fetal calf serum at 37C in a 5% CO2-containing atmosphere.
C2C7 mouse muscle myoblasts were plated at 1 × 104 cells/cm2 and maintained as monolayers in DMEM with 4.5 g/liter glucose, 10% fetal bovine serum, 2 mM glutamine, 200 U/ml penicillin G, and 200 μg/ml streptomycin at 37C in a 5% CO2 atmosphere. Differentiation of C2C7 cells was induced after 2 days in culture by switching the growth medium from one containing 10% fetal bovine serum to one containing 2% heat-inactivated horse serum.
Antiserum
The rabbit polyclonal immune serum against cyclin T2a was produced by immunizing rabbits with a bacterially expressed glutathione-S-transferase (GST)–cyclin T2a fusion protein. The construct comprised the stretch aa 440–663 of the protein. Expression and purification of the fusion protein was done as previously reported (Frangioni and Neel 1993).
In Vitro Transcription-Translation and Immunoprecipitation Assay
One microgram of supercoiled plasmid encoding for human cyclin T2a was used to program a TnT rabbit reticulocyte lysate (Promega; Madison, WI) under the control of the T7 polymerase in the presence of [35S]-methionine. Cell labeling with [35S]-methionine and immunoprecipitation were carried out as previously described (De Luca et al. 1997).
Western Blotting
Western blotting on cell lysates or cell fractions was performed as described previously (Bagella et al. 1998). The anti-cyclin T2a antiserum was used at 1:500 dilution. HSP70 was detected using a mouse monoclonal antibody (HSP01; Oncogene Science, Cambridge, MA).
Immunohistochemistry
Immunohistochemistry was carried out essentially as previously described (De Luca et al. 1997). Briefly, sections from each specimen were cut at 5–7 μm, mounted on glass, and dried overnight at 37C. All sections then were deparaffinized in xylene, rehydrated through a graded alcohol series, and washed in PBS. PBS was used for all subsequent washes and for antibody dilution. Tissue sections were sequentially quenched in 0.5% hydrogen peroxide and were blocked with diluted 10% normal goat anti-rabbit serum (Vector Laboratories; Burlingame, CA). Slides then were incubated at room temperature (RT) for 1 hr with the polyclonal immune serum raised against cyclin T2a at a 1:500 dilution and then with diluted goat anti-rabbit biotinylated antibody (Vector Laboratories) for 30 min. Finally, all slides were processed by the ABC method (Vector Laboratories) for 30 min at RT. Diaminobenzidine was used as the final chromogen and hematoxylin was used as the nuclear counterstain. Negative controls for each tissue section were prepared by substituting the primary antibody with the respective preimmune serum. All samples were processed under the same conditions. Two pathologists (A. Baldi and F. Baldi) evaluated the staining pattern of the proteins separately and scored each specimen for the percentage of positive nuclei identified. An average of 50 fields was observed for each specimen.
Results
Human Cyclin T2a Gene Maps on Chromosome 2q21
Hybridization of the cyclin T2a cDNA probe to human chromosomes showed specific labeling on chromosome 2. Analysis of 22 metaphase spreads revealed that 52% (46/88) of all fluorescence signals hybridized to chromosome 2q. At least one signal specific for chromosome 2q was observed in 20 of the 22 metaphase spreads examined. The distribution of signals was as follows: two chromatids (nine cells), three chromatids (four cells), four chromatids (three cells). The hybridization efficiency was relatively high. All chromosome-specific signals were localized to 2q21 (Figure 1).
Characterization of a New Specific Antiserum for Cyclin T2a
A polyclonal antiserum specific for the human cyclin T2a protein was produced in rabbits. The cDNA for cyclin T2a was translated in vitro and was precipitated using the rabbit polyclonal serum against the GST–cyclin T2a (Figure 2A, Lane c). The polyclonal serum effectively immunoprecipitated in vitro-translated cyclin T2a, whereas undetectable activity was found in the preimmune serum (Figure 2A, Lane b).
The specificity of the antiserum was also assayed by immunoprecipitation. Using cell lysates from [35S]-labeled T98G cells, we observed that cyclin T2a migrated with an apparent molecular mass of approximately 87 kD (Figure 2B, Lane b). The preimmune serum did not show any specific band (Figure 2B, Lane a).
To prove that the antiserum was specific only for cyclin T2a, the in vitro-translated products of both cyclin T1 (Figure 2A, Lane a) and cyclin T2a (Figure 2A, Lane b) were Western blotted and assayed with the antiserum anti-cyclin T2a. As shown in Figure 2C, no crossreaction was observed. An identical specificity was found for the antiserum against cyclin T1, as shown in the previous report (De Luca et al. 2001).
For Western blotting analysis, protein extracts from SAOS-2, T98G, and HeLa cells were employed. Using the specific immune serum, cyclin T2a appeared as a doublet in all the cell lines tested (Figure 2D).
Expression Pattern of Cyclin T2a in Normal Human Tissues
We used immunohistochemical techniques to determine the localization of cyclin T2a in human adult tissues. At least three specimens from different individuals were analyzed for each of the tissues examined. Cyclin T2a protein was found to be ubiquitous, according to mRNA distribution (Peng et al. 1998), but its expression level varied among the different tissues. When the immunohistochemical pattern of expression of cyclin T2a was compared with that specific for cyclin T1 described in the previous article (De Luca et al. 2001), a similar picture was noted. In particular, high expression level of the protein was located in connective cells, in blood and lymphoid tissues, and in pluristratified epithelia. We refer to the previous report on cyclin T1 for a detailed description of the expression pattern in human tissues for cyclin T2a, considering the significant similarities between the expression of the two proteins (De Luca et al. 2001).
Immunohistochemical Expression of Cyclin T2a in Two Cases of Myopathy
Myopathies are non-inflammatory primary degenerative muscle diseases that become symptomatic in infancy or childhood and are characterized by proximal or diffuse muscle weakness (Pearson et al. 1967). Recent evidence suggests that the candidate gene locus for different forms of autosomal dominant myopathies is chromosome 2q21 (Nicolao et al. 1999; Pelin et al. 1999; Xiang et al. 1999).
Our results showed that the human cyclin T2a gene mapped on this chromosomal locus and that cyclin T2a protein level was high in human adult skeletal muscle cells. Moreover, it has already been shown that there is increased cdk9–cyclin T complex activity in some terminally differentiated tissues, and that it may be involved in muscle cell differentiation (Bagella et al. 1998). In effect, preliminary experiments on cyclin T2a during differentiation in C2C7 cells, a mouse myoblast cell line that can be induced to terminally differentiate into myotubes under specific culture conditions (Yaffe and Saxel 1977), demonstrate a positive modulation of cyclin T2a protein levels during differentiation (data not shown).
Drawing on this background, we investigated cyclin T2a immunohistochemical expression in two cases of centronuclear myopathy (Munsat et al. 1969), comparing the results with the expression pattern of cyclin T2a in normal human adult skeletal muscle cells. In the cases of centronuclear myopathy (Figure 3C), expression of the protein was undetectable (Figure 3D), whereas in normal skeletal muscle cells cyclin T2a was expressed at high levels in the nuclei (Figure 3A). In Figure 3B shows a negative control prepared by substituting the primary antiserum with the preimmune serum.
Discussion
A new polyclonal antiserum raised against its protein product was produced to characterize cyclin T2a more accurately. Using this antiserum, we performed a biochemical and immunoistochemical characterization of cyclin T2a. Immunohistochemical analysis of cyclin T2a expression in a large human tissue bank shows that cyclin T2a is widely expressed in all cell types, even if higher levels are found in some terminally differentiated tissues such as muscle, blood, lymphoid tissue, and connective cells. This expression pattern is very similar to that described for cyclin T1 in the previous article (De Luca et al. 2001) and is in agreement with the role of this protein in fundamental cell functions.
Chromosomal localization of cyclin T2a by FISH. Partial human metaphase spread showing specific hybridization signals at chromosome 2q21. The photograph represents computer-generated merged images of red fluorescein signals (arrows) and inverted DAPI-stained chromosomes, the latter displayed in gray. (
Characterization of a new polyclonal antiserum specific for cyclin T2a. In vitro pull-down (
Histology and cyclin T2a immunohistochemistry in muscle fibers and centronuclear muopathy. (
We examined the possible specific role of cyclin T2a in terminally differentiated tissues, such as muscle. It is well known that, during myocyte differentiation, there is downregulation of cyclins D1, E, and A and upregulation of several cdk inhibitors which allow exit from cell cycle and expression of muscle-specific genes (Walsh and Perlman 1997). Differences between cdk9–cyclin T1–T2 complexes and the other cdks–cyclin complexes involved in regulation of the cell cycle may reflect a different function during terminal differentiation in muscle cells. The elevated levels of expression and of kinase activity of cdk9, together with the fact that cdk9–cyclin T1–T2 complexes are not cell cycle-regulated, distinguish cdk9 from the other cdks (MacLachlan et al. 1995; De Falco and Giordano 1998). In this context, the induction of myocyte differentiation on partial serum withdrawal, as occurs with C2C7 cells, is a good in vitro system in which to look at the cdk9–cyclin T1–T2 functions during terminal differentiation (Bagella et al. 1998). The high immunohistochemical expression level of cyclin T2a in adult muscle cells, as well as the peak of expression reached during in vitro muscle differentiation (data not shown), suggests that cdk9–cyclin T2a complex might promote myogenic differentiation, eliciting the expression of muscle-specific genes.
Because cyclin T2a gene was mapped on chromosome 2q21, a locus recently linked to different forms of myopathy (Nicolao et al. 1999; Pelin et al. 1999; Xiang et al. 1999), we looked at the immunohistochemical expression of cyclin T2a in tissue from two patients with centronuclear myopathy. In this form of myopathy, most of the myofibers contain central nuclei, which give them an appearance resembling embryonal myotubules (Munsat et al. 1969). Whereas the immunohistochemical expression of cyclin T2a is very high in normal skeletal muscle cells, its expression is undetectable in the tissue from two patients with centronuclear myopathy.
We conclude that, apart from its involvement in transcription regulation, the cdk9–cyclin T2a complex could have other functions. On the basis of high expression level in differentiated muscle tissues, we speculate that cdk9–cyclin T2a activity could be involved in myogenic differentiation. The observation that immunohistochemical expression of cyclin T2a was un-detectable in two patients with centronuclear myopathy suggests an involvement of the cdk–9cyclin T2a complex in this disease. This is in agreement with the finding that the gene maps on a locus linked with different forms of myopathy and with its putative role in myogenic differentiation.
Footnotes
Acknowledgements
Supported in part by AIRC and Ministero della Sanità grants to M.G.P., by a Second University of Naples and a CNR grant to A.D.L., by NIH grants RO1 CA 60999-01A1 and PO1 NS 36466 to A.G., by NCI grants CA45745 and CA06927 (Fox Chase Cancer Center), and by an appropriation from the Commonwealth of Pennsylvania to J.R.T. A.B. is recipient of an FIRC fellowship.
We thank Mr Tullio Battista (Regina Elena Institute; Rome, Italy) for his excellent technical assistance and Dr J.J. Gartland (Thomas Jefferson University; Philadelphia, PA) for editing the manuscript.