Ribonuclease L Proteolysis in Peripheral Blood Mononuclear Cells of Chronic Fatigue Syndrome Patients
2002; Elsevier BV; Volume: 277; Issue: 38 Linguagem: Inglês
10.1074/jbc.m201263200
ISSN1083-351X
AutoresEdith Demettre, Lionel Bastide, A. D'haese, Karen De Smet, Kenny De Meirleir, K.P. Tiev, Patrick Englebienne, Bernard Lebleu,
Tópico(s)Fibromyalgia and Chronic Fatigue Syndrome Research
ResumoA 37-kDa binding polypeptide accumulates in peripheral blood mononuclear cell (PBMC) extracts from chronic fatigue syndrome (CFS) patients and is being considered as a potential diagnostic marker (De Meirleir, K., Bisbal, C., Campine, I., De Becker, P., Salehzada, T., Demettre, E., and Lebleu, B. (2000) Am. J. Med. 108, 99–105). We establish here that this low molecular weight 2-5A-binding polypeptide is a truncated form of the native 2-5A-dependent ribonuclease L (RNase L), generated by an increased proteolytic activity in CFS PBMC extracts. RNase L proteolysis in CFS PBMC extracts can be mimicked in a model system in which recombinant RNase L is treated with human leukocyte elastase. RNase L proteolysis leads to the accumulation of two major fragments with molecular masses of 37 and 30 kDa. The 37-kDa fragment includes the 2-5A binding site and the N-terminal end of native RNase L. The 30-kDa fragment includes the catalytic site in the C-terminal part of RNase L. Interestingly, RNase L remains active and 2-5A-dependent when degraded into its 30- and 37-kDa fragments by proteases of CFS PBMC extract or by purified human leukocyte elastase. The 2-5A-dependent nuclease activity of the truncated RNase L could result from the association of these digestion products, as suggested in pull down experiments. A 37-kDa binding polypeptide accumulates in peripheral blood mononuclear cell (PBMC) extracts from chronic fatigue syndrome (CFS) patients and is being considered as a potential diagnostic marker (De Meirleir, K., Bisbal, C., Campine, I., De Becker, P., Salehzada, T., Demettre, E., and Lebleu, B. (2000) Am. J. Med. 108, 99–105). We establish here that this low molecular weight 2-5A-binding polypeptide is a truncated form of the native 2-5A-dependent ribonuclease L (RNase L), generated by an increased proteolytic activity in CFS PBMC extracts. RNase L proteolysis in CFS PBMC extracts can be mimicked in a model system in which recombinant RNase L is treated with human leukocyte elastase. RNase L proteolysis leads to the accumulation of two major fragments with molecular masses of 37 and 30 kDa. The 37-kDa fragment includes the 2-5A binding site and the N-terminal end of native RNase L. The 30-kDa fragment includes the catalytic site in the C-terminal part of RNase L. Interestingly, RNase L remains active and 2-5A-dependent when degraded into its 30- and 37-kDa fragments by proteases of CFS PBMC extract or by purified human leukocyte elastase. The 2-5A-dependent nuclease activity of the truncated RNase L could result from the association of these digestion products, as suggested in pull down experiments. Chronic fatigue syndrome is characterized by long-lasting and debilitating fatigue, myalgia, impairment of neurocognitive functions, and flu-like symptoms, which are often severely worsened after physical exercise. A case definition has been proposed by the Center for Disease Control in Atlanta under the name of Holmes (2Holmes G.P. Kaplan J.E. Gantz N.M. Komaroff A.L. Schonberger L.B. Straus S.E. Jones J.F. Dubois R.E. Cunningham-Rundles C. Pahwa S. Ann. Intern. Med. 1988; 108: 387-389Crossref PubMed Scopus (1447) Google Scholar) and Fukuda (3Fukuda K. Straus S.E. Hickie I. Sharpe M.C. Dobbins J.G. Komaroff A. Ann. Intern. Med. 1994; 121: 953-959Crossref PubMed Scopus (4310) Google Scholar) criteria. Most of these symptoms are unfortunately common to other diseases, thus complicating a diagnosis that still relies on extensive clinical testing to exclude other pathologies (4Komaroff A.L. Buchwald D.S. Annu. Rev. Med. 1998; 49: 1-13Crossref PubMed Scopus (132) Google Scholar). A large proportion of the patients report an infectious episode at the onset of their chronic fatigue. No single agent has been conclusively associated with the disease, although several candidates including human T-cell lymphotrophic virus-1, human herpesvirus-6,Enterovirus, or mycoplasma have been proposed (5DeFreitas E. Hilliard B. Cheney P.R. Bell D.S. Kiggundu E. Sankey D. Wroblewska Z. Palladino M. Woodward J.P. Koprowski H. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 2922-2926Crossref PubMed Scopus (130) Google Scholar, 6Ablashi D.V. Eastman H.B. Owen C.B. Roman M.M. Friedman J. Zabriskie J.B. Peterson D.L. Pearson G.R. Whitman J.E. J. Clin. Virol. 2000; 16: 179-191Crossref PubMed Scopus (153) Google Scholar, 7Swanink C.M. Melchers W.J. van der Meer J.W. Vercoulen J.H. Bleijenberg G. Fennis J.F. Galama J.M. Clin. Infect. Dis. 1994; 19: 860-864Crossref PubMed Scopus (48) Google Scholar, 8Nasralla M. Haier J. Nicolson G.L. Eur. J. Clin. Microbiol. Infect. Dis. 1999; 18: 859-865Crossref PubMed Scopus (87) Google Scholar). Dysregulation of immune functions has also been suggested and natural killer cell cytotoxicity was significantly diminished in patients (when compared with normal controls) (9Klimas N.G. Salvato F.R. Morgan R. Fletcher M.A. J. Clin. Microbiol. 1990; 28: 1403-1410Crossref PubMed Google Scholar). These observations have prompted studies of possible dysfunctioning of interferon-induced responses. An up-regulation of the 2-5A/RNase L antiviral pathway in peripheral blood mononuclear cells (PBMC) 1The abbreviations used are: PBMC, peripheral blood mononuclear cell; CFS, chronic fatigue syndrome; Ni-NTA, nickel nitrilotriacetic acid; HLE, human leukocyte elastase; 2-5A, 2′,5′-linked oligoadenylate; RNase L, 2-5A-dependent ribonuclease L; pCp, cytidine 3′,5′- bisphosphate. of CFS patients has been described (10Suhadolnik R.J. Reichenbach N.L. Hitzges P. Sobol R.W. Peterson D.L. Henry B. Ablashi D.V. Muller W.E. Schroder H.C. Carter W.A. Strayer D., R. Clin. Infect. Dis. 1994; 18 Suppl. 1: 96-104Crossref Scopus (119) Google Scholar). RNase L is the terminal enzyme in the 2-5A synthetase/RNase L antiviral pathway and plays an essential role in the elimination of viral mRNAs (for review, see Ref. 11Bastide L. Demettre E. Martinand-Mari C. Lebleu B. Englebienne P. De Meirleir K. Chronic Fatigue Syndrome: A Biological Approch. CRC Press, Inc., Boca Raton, FL2002: 1-15Crossref Google Scholar). Activation of RNase L requires the binding of a small 2′,5′-linked oligoadenylate (2-5A). Intriguingly, a low molecular weight 2-5A-binding polypeptide has been observed in a subset of patients diagnosed for CFS (12Suhadolnik R.J. Peterson D.L. O'Brien K. Cheney P.R. Herst C.V. Reichenbach N.L. Kon N. Horvath S.E. Iacono K.T. Adelson M.E., De Meirleir K., De Becker P. Charubala R. Pfleiderer W. J. Interferon Cytokine Res. 1997; 17: 377-385Crossref PubMed Scopus (105) Google Scholar). Similar observations have been made in a larger study including CFS patients and control groups (1De Meirleir K. Bisbal C. Campine I., De Becker P. Salehzada T. Demettre E. Lebleu B. Am. J. Med. 2000; 108: 99-105Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar). In the latter study the presence of a 37-kDa 2-5A-binding polypeptide was found in a significant proportion of the CFS population but absent in most controls, thus providing the basis for a potential biochemical marker of CFS. This low molecular weight 2-5A-binding polypeptide is related to RNase L since it is recognized by polyclonal antibodies raised against RNase L (12Suhadolnik R.J. Peterson D.L. O'Brien K. Cheney P.R. Herst C.V. Reichenbach N.L. Kon N. Horvath S.E. Iacono K.T. Adelson M.E., De Meirleir K., De Becker P. Charubala R. Pfleiderer W. J. Interferon Cytokine Res. 1997; 17: 377-385Crossref PubMed Scopus (105) Google Scholar). Moreover, an increased nuclease activity has been reported in a subset of CFS patients whose PBMC extracts did not contain any full-size 83-kDa RNase L (13Shetzline S.E. Suhadolnik R.J. J. Biol. Chem. 2001; 276: 23707-23711Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar). This observation has remained puzzling since the P-loop region (which is responsible for 2-5A binding) and the C-terminal end (which contains the catalytic site of RNase L) are far apart and cannot be fitted within a single 37-kDa polypeptide. In the present study, we establish that an increased proteolytic activity in PBMC extracts from CFS patients is responsible for the accumulation of this truncated form of RNase L and that a 2-5A-dependent nucleolytic activity is maintained in the absence of intact RNase L. Moreover, we demonstrate that proteolysis of RNase L by proteases in CFS PBMC extract or by purified human leukocyte elastase (HLE) gives rise to 2 major fragments with molecular masses of 37 and 30 kDa. The 37-kDa fragment includes the 2-5A binding site and the N-terminal end of full-size RNase L. The 30-kDa fragment starts in the second half of the protein kinase homology domain and includes the catalytic site in the C-terminal part of RNase L. Recombinant RNase L was cloned and expressed in baculovirus-infected Sf21 insect cells by ATG Laboratories (Eden Prairie, MN). A His6 tag was inserted at the N terminus of the protein, which was purified to 90–95% homogeneity by metal chelate chromatography on nickel nitrilotriacetic acid (Ni-NTA)-agarose. Venous blood samples were drawn from patients who fulfilled the Holmes (2Holmes G.P. Kaplan J.E. Gantz N.M. Komaroff A.L. Schonberger L.B. Straus S.E. Jones J.F. Dubois R.E. Cunningham-Rundles C. Pahwa S. Ann. Intern. Med. 1988; 108: 387-389Crossref PubMed Scopus (1447) Google Scholar) and Fukuda (3Fukuda K. Straus S.E. Hickie I. Sharpe M.C. Dobbins J.G. Komaroff A. Ann. Intern. Med. 1994; 121: 953-959Crossref PubMed Scopus (4310) Google Scholar) criteria for CFS or from healthy individuals. Patients with CFS and healthy individuals were recruited from the Free University of Brussels (Brussels, Belgium) and from the Saint Antoine Hospital (Paris, France). The procedure for isolating PBMC was started within 2 h of sampling. Heparinized whole blood was diluted 1:1 with phosphate-buffered saline. Two volumes of diluted blood were overlaid on 1 volume of Ficoll Histopaque® (Sigma-Aldrich) (density of 1.080 g/ml) and centrifuged at 20 °C at 500 × g for 30 min. The PBMC layer was removed, washed with phosphate-buffered saline, and centrifuged. The isolated PBMC pellets were resuspended in 5 ml of red blood cell lysis buffer (155 mm NH4Cl, 10 mmNaHCO3, pH 7.4, 0.1 mm EDTA), kept on ice for 5 min, and centrifuged (20 °C, 500 × g, 10 min). The PBMC were washed with phosphate-buffered saline and centrifuged again. The pellets were frozen at −80 °C until use. PBMC from six CFS patients and from three healthy individuals were pooled to create, respectively, CFS PBMC pellets and control PBMC pellets. The cells were resuspended in 2 volumes of hypotonic buffer (20 mm Hepes, pH 7.5, 10 mm potassium acetate, 1.5 mm magnesium acetate, 0.5% (v/v) ethylphenylpolyethyleneglycol (Nonidet P 40) detergent) by repeated pipetting and centrifuged at 10,000 × g for 15 min at 4 °C. The protein concentration in the supernatant was determined by spectrophotometry (14Whitaker J.R. Granum P.E. Anal. Biochem. 1980; 109: 156-159Crossref PubMed Scopus (491) Google Scholar). Chemically synthesized 5′-monophosphorylated 2′,5′-linked oligoadenylate tetramer (2-5A) was generously given by Prof. W. Pfleiderer (University of Konstanz, Konstanz, Germany). The 2-5A probe was labeled by ligation of [32P]pCp (specific activity, 3,000 Ci/mmol, ICN Pharmaceuticals) to the 3′-terminus of 2-5A with T4 RNA ligase (Amersham Biosciences) and purified by high performance liquid chromatography on a Hypersil ODS 5-μm C18 column. The terminal 3′-phosphate group was removed by treatment with T4 polynucleotide kinase (Invitrogen) thanks to its 3′-phosphorylase activity. The pH was adjusted to 4.7, the 3′-ribose residue was oxidized with 10 mm sodium metaperiodate, and the pH was readjusted to 8. The 3′-oxidized 32P-labeled 2-5ApC (2-5A probe) (3,000 Ci/mmol) was incubated with PBMC extract or with recombinant RNase L for 30 min at 4 °C and for a further 30 min at 4 °C with 20 mm NaBH3CN. The polypeptides were fractionated by 12% (w/v) SDS-PAGE (15Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (212360) Google Scholar) and detected by autoradiography. Control or CFS PBMC extracts (50 μg of total proteins) were covalently labeled with the 2-5A probe as described previously. Extracts were incubated with or without 10 mm HLE inhibitor III (Calbiochem) and diluted in Me2SO in 10 μl of buffer A (20 mm Tris-HCl, pH 7.5, 10 mm MgCl2, 8 mm2-mercaptoethanol, 90 mm KCl, and 0.1 mm ATP) at 37 °C. At the indicated time, extracts were heated to 95 °C in gel-loading buffer for 3 min. The proteins were then separated on SDS-PAGE and detected by autoradiography. When HLE inhibitor III was used, a control was added with Me2SO alone to rule out an effect of Me2SO on HLE activity. The recombinant RNase L (250 ng) was covalently labeled with the 2-5A probe and incubated for the indicated time in 10 μl of buffer A at 37 °C with control or CFS PBMC extracts (2.5 μg of total proteins) or with 4 × 10−3 units of HLE (Sigma-Aldrich) in the presence or absence of 10 mm HLE inhibitor III. The proteins were then separated on SDS-PAGE and detected by autoradiography. The extract from Daudi lymphoblastoid cells (50 μg of total proteins) was covalently labeled with the 2-5A probe and incubated for 30 min at 37 °C in 10 μl of buffer A with HLE. The proteins were then separated on SDS-PAGE and detected by autoradiography. The assay of RNase L nuclease activity was adapted from Carroll et al. (16Carroll S.S. Chen E. Viscount T. Geib J. Sardana M.K. Gehman J. Kuo L.C. J. Biol. Chem. 1996; 271: 4988-4992Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar) and Dong and Silverman (17Dong B. Silverman R.H. J. Biol. Chem. 1997; 272: 22236-22242Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Briefly, the RNase L oligonucleotide substrate C11U2C7 (Xeragon) was labeled with [32P]pCp (5 mCi/ml) with T4 RNA ligase (AmershamBiosciences) in 50 mm Hepes, pH 7.5, 15 mmMgCl2, 2 mm dithiothreitol, 5 μg/ml bovine serum albumin, and 25 μm ATP for 15 min at 30 °C. Recombinant RNase L (400 ng) was digested, or not, by HLE (5 × 10−3 units) or by a low concentration of PBMC CFS extract (200 ng of total proteins) and then incubated with or without 100 nm of 2-5A for 10 min at 4 °C. Aliquots from this reaction mixture were covalently labeled with the 2-5A probe, and the amounts of 83- and 37-kDa RNase L were quantified after SDS-PAGE analysis and autoradiography. For measurement of RNase L activity, aliquots of the reaction mixture were supplemented with 37.5 pmol (including 10% of [32P]C11U2C7pCp) of RNase L substrate in 10 μl of buffer A and further incubated for 30 min at 30 °C. Aliquots were taken again to quantify the amounts of 83- and 37-kDa RNase L after the activity incubation. The cleavage products of [32P]C11U2C7pCp were separated on a 20% (w/v) acrylamide, 7 m urea gel in Tris acetate EDTA and analyzed by autoradiography. Recombinant RNase L (10 μg) were digested, or not, by HLE (0.2 units) or by minute amounts of PBMC CFS extract (5 μg of total proteins) in buffer A. After analysis by SDS-PAGE, proteins were revealed by a zinc stain kit for electrophoresis (Bio-Rad) or transferred by Western blot onto a polyvinylidene difluoride membrane and stained with Ponceau red. The major 30- and 37-kDa bands were then excised from the Western blot to be analyzed on a Procise® protein sequencer (Applied Biosystem). Recombinant RNase L (1 μg) was digested, or not, by HLE (0.02 units) or by a low concentration of PBMC CFS extract (0.5 μg of total proteins) in buffer A. After separation by SDS-PAGE and transfer to a polyvinylidene difluoride membrane, the His6-tag of the recombinant RNase L was detected with Ni-NTA horseradish peroxidase conjugate (Qiagen). Recombinant RNase L was incubated overnight at 4 °C with Ni-NTA magnetic-agarose beads (Qiagen) in 50 mm NaH2PO4, pH 8, 300 mm NaCl, 20 mm imidazole, and 50% (v/v) glycerol. The beads were then washed twice with 50 mmNaH2PO4, pH 8, 300 mm NaCl, and 20 mm imidazole to eliminate unfixed proteins. An aliquot of RNase L bound to the beads was heated to 95 °C in gel loading buffer for 3 min. The other part (1 μg) was digested by HLE (10−2 units) for 15 min at 30 °C in phosphate-buffered saline, washed twice, and heated to 95 °C in gel-loading buffer for 3 min. His6-tagged proteins were then fractionated by SDS-PAGE and detected by a zinc stain kit for electrophoresis (Bio-Rad). In our initial studies (1De Meirleir K. Bisbal C. Campine I., De Becker P. Salehzada T. Demettre E. Lebleu B. Am. J. Med. 2000; 108: 99-105Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar), we established that PBMC extracts from CFS patients were characterized by the presence of a low molecular mass (37 kDa) 2-5A-binding polypeptide, whereas an 83-kDa band was predominant in extracts from control PBMC. In these experiments, several protease inhibitors were included at the time of PBMC lysis to limit breakdown of native 83-kDa RNase L during the preparation and the processing of the PBMC extracts. An essential concern was the link between this low molecular mass (37 kDa) 2-5A-binding polypeptide and native RNase L. At this stage, we could not establish whether this 37-kDa 2-5A-binding polypeptide was a new protein, distinct from native RNase L, or was processed from RNase L by proteolysis in intact cells. To discriminate between these two possibilities, PBMC extracts were prepared in the absence of protease inhibitors from healthy individuals or from CFS patients (as described under "Experimental Procedures") and incubated at 37 °C for increasing periods of time. The 83-kDa RNase L rapidly disappeared upon incubation in PBMC CFS extract, with the concomitant accumulation of a 2-5A probe-labeled band migrating with an apparent molecular mass of 37 kDa (Fig.1B). Additional minor 2-5A probe-labeled polypeptides were detected as well, which probably represent unstable intermediates in the processing of RNase L. In contrast, the 83-kDa RNase L was much more stable in PBMC control extract (Fig.1 A). To further demonstrate an increased proteolytic activity in CFS extracts, recombinant RNase L was covalently labeled with the 2-5A probe and incubated with a low amount of unlabeled control or CFS PBMC extract (Fig. 2). Recombinant RNase L was rapidly converted into a 37-kDa truncated RNase L when incubated with CFS extract and remained essentially stable in the presence of control extract, in keeping with an increased proteolytic activity in the former one. Fingerprint analysis, by limited proteolysis with chymotrypsin or V8 protease, revealed that the 37-kDa band arising from the digestion of the 2-5A probe-labeled recombinant RNase L with CFS extracts gave rise to a proteolytic signature that was identical to that of the 37-kDa-labeled material found in PBMC extracts of CFS patients (data not shown). The biochemical characterization of RNase L degradation products in human PBMC extracts from CFS patients was difficult because of the low abundance of RNase L. We have therefore attempted to degrade recombinant RNase L by purified proteases. To validate this model, degradation of recombinant RNase L by CFS extract was used as control. Preliminary studies have established that several proteases known to be present in PBMC were able to degrade RNase L and to lead to the accumulation of a 37-kDa truncated form (data not shown). HLE appeared particularly interesting in this respect since HLE inhibitor III, a specific peptidic HLE inhibitor, partially inhibited the degradation of endogenous RNase L in PBMC CFS extract (Fig.3). Along the same line, endogenous RNase L in a Daudi lymphoblastoid cell extract (Fig.4 A) or recombinant RNase L (Fig. 4 B) was degraded to a 37-kDa truncated RNase L by purified HLE (Fig. 4, A (lanes 2–4) andB (lane 2)). It should be noted that the degradation of recombinant RNase L by PBMC CFS extract (Fig.4 B, lane 5) gave rise to an identical pattern. Moreover, the addition of HLE inhibitor III blocked the degradation of recombinant RNase L by purified HLE (Fig. 4 B, lanes 3 and 4) or by PBMC CFS extracts (lanes 6and 7).Figure 4RNase L is proteolyzed into a 37-kDa truncated form by purified HLE. 2-5A-labeled endogenous RNase L in a Daudi lymphoblastoid cell extract (panel A) was incubated 30 min at 37 °C in the absence (lane 1) or the presence (lanes 2–4) of increasing amounts (5 × 10−4, 5 × 10−3, and 5 × 10−2 units, respectively) of purified HLE. 2-5A-labeled recombinant RNase L (panel B) was incubated 15 min at 37 °C in the absence of proteases (lane 1), the presence of 4 × 10−3 HLE units (lanes 2–4), or the presence of minute amounts of CFS extract (lanes 5–7) as a source of proteases. HLE inhibitor III was eventually added (lanes 3 and 6). Controls for the absence of Me2SO (DMSO) effects are shown in lanes 4 and 7.View Large Image Figure ViewerDownload (PPT) Evaluating the RNase L-associated biological activity in unfractionated cell extracts was difficult due to its low abundance, the presence of inhibitors, and the presence of other nucleases. We therefore made use of the model system described above and of the short radiolabeled RNA substrate (C11U2C7) described by Carrollet al. (16Carroll S.S. Chen E. Viscount T. Geib J. Sardana M.K. Gehman J. Kuo L.C. J. Biol. Chem. 1996; 271: 4988-4992Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar). Recombinant RNase L was digested by purified HLE or by a low concentration of PBMC CFS extract, as described under "Experimental Procedures," before being analyzed for nucleolytic activity. As seen in Fig. 5, the [32P]C11U2C7pCp substrate was cleaved to the expected 8-mer labeled degradation product ([32P]C7pCp) even when the recombinant RNase L had been degraded by purified HLE or by a low concentration of CFS extract, as a source of proteases. It should be noted that the amount of CFS extract (200 ng of total protein) used as a source of proteases was not sufficient in its own right (no recombinant RNase L added) to degrade labeled C11U2C7 (Fig. 5,lanes 7 and 8). The nucleolytic activity observed was strictly 2-5A-dependent for degraded RNase L as for intact RNase L (Fig. 5, lanes 3–6). In previous results sections the proteolysis of recombinant or endogenous RNase L was monitored by 2-5A probe labeling. This strategy was advantageous in being highly sensitive and specific for RNase L. It, of course, did not allow the detection of unlabeled proteolytic degradation products. We therefore made use of a sensitive zinc-imidazole staining method to detect proteolysis products of recombinant RNase L. As shown in Fig.6, recombinant RNase L migrated as a major 83-kDa polypeptide as expected. Upon incubation with purified HLE, RNase L was rapidly degraded into two groups of fragments migrating with molecular masses around 30 and 37 kDa, respectively. The latter had a similar electrophoretic mobility as the 2-5A labeled 37-kDa truncated form of RNase L (data not shown and Fig. 4). The same bands were detected when CFS extract was used as a source of endogenous proteases, although the pattern was complicated by the presence of PBMC extract proteins. As an example, the major 42-kDa band corresponded to β-actin. The main cleavage fragments (as indicated by the arrows in Fig. 6) were collected after electrophoretic blotting and processed for microsequencing of their N-terminal ends. The N-terminal sequences of the 30-kDa fragments were500HLADFDKSI508 and492LIDSKKAAH500 when produced by HLE or by CFS extract protease digestion, respectively. The observed electrophoretic mobilities of these 30-kDa fragments were in agreement with the calculated mass for peptides extending from the proteolytic cleavage site to the C-terminal end of RNase L. No N-terminal sequence could be identified for the 37-kDa fragments produced by HLA or CFS extract proteolysis of recombinant RNase L, suggesting an N-terminally blocked end. This was expected if these 37-kDa fragments started at the N-terminal end of a recombinant protein produced in a baculovirus expression system. It was indeed verified that the full-size recombinant RNase L had a blocked N-terminal end (data not shown). Moreover, recombinant RNase L (which is tagged with His6 at its N-terminal end) was digested with HLE or with CFS extract, and the His6 tag in the digestion products was identified with a specific Ni-NTA horseradish peroxidase conjugate. As shown in Fig. 7, an 83-kDa band was labeled in intact recombinant RNase L, and a major 37-kDa band was detected after proteolysis. These major 30- and 37 kDa cleavage fragments remain associated, as evidenced by pull down experiments. We took advantage of the presence of the His6 tag at the N-terminal end of recombinant RNase L to immobilize the enzyme on Ni-NTA magnetic-agarose beads (Fig.8, lane 1). Treatment of bound RNase L with HLE followed by extensive washings led to proteolysis of the 83-kDa material into the expected 30- and 37-kDa fragments, which both remained bound to the beads (Fig. 8, lane 3). None of the 30-kDa material was released in the supernatant after proteolysis (Fig. 8, lane 2). We have established here that the 37-kDa 2-5A-binding polypeptide that accumulates in CFS PBMC can be generated by proteolysis of endogenous RNase L (Fig. 1) or by the incubation of recombinant RNase L with a low concentration of PBMC CFS extract (Fig. 2), in keeping with preliminary studies from our group (18Lebleu B. Herst C.V. Bastide L. Demettre E. De Smet K. D'Haese A. Roelens S. De Meirleir K. Le Roy F. Silhol M. Englebienne P. Eur. Cytokine Netw. 2000; 11 (abstr.): 101Google Scholar, 19Roelens S. Herst C.V. D'Haese A., De Smet K. Frémont M., De Meirleir K. Englebienne P. J. Chronic Fatigue Syndr. 2001; 8: 63-82Crossref Scopus (9) Google Scholar). Increased proteolytic activity in CFS PBMC, thus, appears to be the major cause for the occurrence of this 37-kDa 2-5A-binding polypeptide and rules out differential RNA splicing or protein splicing as discussed by Shetzline et al.(20Shetzline S.E. Martinand-Mari C. Reichenbach N.L. Buletic Z. Lebleu B. Pfleiderer W. Charubala R., De Meirleir K., De Becker P. Peterson D.L. Herst C.V. Englebienne P. Suhadolnik R.J. J. Interferon Cytokine Res. 2002; 22: 443-456Crossref PubMed Scopus (25) Google Scholar). Proteases play an important role in numerous physiological responses and, in particular, in inflammation and apoptosis. Some of the symptoms observed in CFS can be related to a dysregulation of these functions. As an example, CFS is associated with increased IL6 secretion and elevated α2-macroglobulin, suggesting an acute phase inflammatory response (21Cannon J.G. Angel J.B. Ball R.W. Abad L.W. Fagioli L. Komaroff A.L. J. Clin. Immunol. 1999; 19: 414-421Crossref PubMed Scopus (67) Google Scholar). Moreover, infections by various pathogenic agents (5DeFreitas E. Hilliard B. Cheney P.R. Bell D.S. Kiggundu E. Sankey D. Wroblewska Z. Palladino M. Woodward J.P. Koprowski H. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 2922-2926Crossref PubMed Scopus (130) Google Scholar, 6Ablashi D.V. Eastman H.B. Owen C.B. Roman M.M. Friedman J. Zabriskie J.B. Peterson D.L. Pearson G.R. Whitman J.E. J. Clin. Virol. 2000; 16: 179-191Crossref PubMed Scopus (153) Google Scholar, 7Swanink C.M. Melchers W.J. van der Meer J.W. Vercoulen J.H. Bleijenberg G. Fennis J.F. Galama J.M. Clin. Infect. Dis. 1994; 19: 860-864Crossref PubMed Scopus (48) Google Scholar, 8Nasralla M. Haier J. Nicolson G.L. Eur. J. Clin. Microbiol. Infect. Dis. 1999; 18: 859-865Crossref PubMed Scopus (87) Google Scholar) have been described in CFS patients and can lead to inflammatory processes. Finally, an increased apoptosis has been documented in CFS PBMC (22Vojdani A. Ghoneum M. Choppa P.C. Magtoto L. Lapp C.W. J. Intern. Med. 1997; 242: 465-478Crossref PubMed Scopus (47) Google Scholar), and RNase L is known to be involved in apoptosis in response to various stimuli (23Zhou A. Paranjape J. Brown T.L. Nie H. Naik S. Dong B. Chang A. Trapp B. Fairchild R. Colmenares C. Silverman R.H. EMBO J. 1997; 16: 6355-6363Crossref PubMed Scopus (460) Google Scholar, 24Rusch L. Zhou A. Silverman R.H. J. Interferon Cytokine Res. 2000; 20: 1091-1100Crossref PubMed Scopus (79) Google Scholar). A panel of proteases known to act in cell apoptosis or in inflammatory responses have, therefore, been evaluated for their abilities to degrade RNase L. Caspase 3 is not activated in CFS PBMC extracts, 2E. Demettre and C. Bisbal, unpublished observations. and no cleavage of recombinant RNase L by caspase 3 (25Roelens, S., Herst, C. V., De Smet, K., D'Haese, A., Frémont, M., De Meirleir, K., Englebienne, P., American Association for Chronic Fatigue Syndrome Fifth International Research, Clinical and Patient Conference, January 26–29, 2001, AACFS, 2001, 70, AACFS, Seattle, WA.Google Scholar) has been observed. HLE and cathepsin G belong to the neutral serine protease family from azurophilic granules of myeloid leukocytes and are involved in host defenses against pathogens and inflammatory responses. Interestingly, the incubation of recombinant RNase L with cathepsin G, HLE, or calpain give rise to cleavage products that are similar to those found upon PBMC CFS incubation (Fig.1). 3E. Demettre, L. Bastide, and A. D'Haese, unpublished observations. This criterion alone, however, does not implicate these proteases in the processing of RNase L. Indeed, protein cleavage by a protease at a given site often reflects protein structure only (26Fontana A. Polverino de Laureto P., De Filippis V. Scaramella E. Zambonin M. Fold. Des. 1997; 2: 17-26Abstract Full Text Full Text PDF PubMed Scopus (286) Google Scholar). An additional argument for HLE involvement in the proteolysis of RNase L in CFS PBMC extract is the observation that HLE inhibitor III, a specific peptide inhibitor of HLE (27Powers J.C. Gupton B.F. Harley A.D. Nishino N. Whitley R.J. Biochim. Biophys. Acta. 1977; 485: 156-166Crossref PubMed Scopus (175) Google Scholar, 28Stein R.L. Trainor D.A. Biochemistry. 1986; 25: 5414-5419Crossref PubMed Scopus (54) Google Scholar), inhibits, at least, in part the accumulation of the 37-kDa truncated RNase L in these extracts (Fig. 3). Although the role of other proteases cannot be excluded at this stage, enhanced HLE activity appears to be involved in the increased proteolysis of RNase L in CFS PBMC. Whether this increased proteolytic activity results from enzyme activation, enzyme redistribution, or other mechanisms is presently unknown. Whatever the underlying mechanism, degradation of recombinant RNase L by purified HLE represents a convenient tool to generate RNase L cleavage fragments for further biochemical characterization. Indeed, the cleavage of recombinant RNase L by HLE mimics RNase L proteolysis in CFS PBMC by the following criteria. First, the same products are generated upon cleavage of recombinant RNase L by purified HLE or by CFS extract, whether 2-5A probe labeling (Fig. 4 B), zinc-imidazole staining (Fig. 6), or detection of the N-terminal His tag (Fig. 7) are monitored. Second, the 2-5A-labeled 37-kDa fragments produced by the proteolysis of endogenous RNase L in a CFS PBMC extract or by the degradation of recombinant RNase L with HLE give rise to identical sets of 2-5A-labeled peptides after limited proteolysis with chymotrypsin or with V8 protease (data not shown). Third, the incubation of recombinant RNase L with either HLE or CFS extracts gives rise to closely related 30-kDa fragments, as indicated by the microsequencing of their N-terminal sequences (Fig.9). Finally, the nucleolytic activity of both sets of cleavage products is similar (Fig. 5). The two major polypeptides produced by the proteolysis of RNase L have molecular masses of 37 and 30 kDa. The 37-kDa fragment includes the N-terminal end (since it contains the N-terminal His6 tag) and the 2-5A binding site (which has been allocated to the P-loop motifs in ankyrin domains 7 and 8) (29Silverman R.H. D'Alessio G. Riordan J.F. Ribonucleases: Structures and Functions. Academic Press, Inc., New York1997: 515-551Crossref Google Scholar) (Fig.9). Its C-terminal end has not been precisely located, but it most probably ends shortly after the P-loops at the end of the ankyrin domains region (Fig. 9). The 30-kDa fragment starts in the second half of the protein kinase homology domain (at amino acid 500 after HLE proteolysis) and extends until the C-terminal end of RNase L, thus encompassing its catalytic site (Fig. 9). Proteolysis of RNase L then removes the first half of the protein kinase homology domain. Whether the increased proteolysis of RNase L in CFS PBMC has a significant qualitative or quantitative effect on its biological activity has not been fully appreciated. Shetzline and Suhadolnik (13Shetzline S.E. Suhadolnik R.J. J. Biol. Chem. 2001; 276: 23707-23711Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar) document a slightly increased nuclease activity in affinity-purified CFS PBMC extracts. Unfortunately, their experimental model could not exclude a contribution of other nucleases in the degradation of the poly(U) substrate used in their experiments. We made use of the simplified biological model of proteolysis described above and of a short synthetic C11U2C7-labeled substrate containing an unique and specific cleavage site for RNase L. Its 2-5A-dependent cleavage by RNase L at this site gives rise to a single 8-mer cleavage fragment that can easily be resolved and quantified by gel electrophoresis. This assay confirms that the biological activity of proteolyzed RNase L is not dramatically different from native RNase L, in agreement with Shetzline and Suhadolnik (13Shetzline S.E. Suhadolnik R.J. J. Biol. Chem. 2001; 276: 23707-23711Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar). A 2-5A-dependent nuclease activity remains after the proteolysis of recombinant RNase L in its 30- and 37-kDa major cleavage fragments. The 30-kDa fragment, which encompasses the C-terminal catalytic site of RNase L, could potentially be active but does not include the 2-5A binding site. Interestingly in this respect, an N-terminal-deleted recombinant form of RNase L lacking all ankyrin domains (which could be analogous to the 30-kDa fragment described here) exhibits a 2-5A-independent nuclease activity (17Dong B. Silverman R.H. J. Biol. Chem. 1997; 272: 22236-22242Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). On the other hand, the 37-kDa fragment, which encompasses the His6-tagged N-terminal end of RNase L and the 2-5A binding site, cannot include the catalytic site. In principle, therefore, a 2-5A-dependent nuclease activity cannot be assigned to the 30-kDa or to the 37-kDa fragment alone. However, these two fragments could remain linked together by disulfide bridges or by non-covalent bonds after proteolysis, as suggested by the pull down experiments (Fig. 8), and lead to the observed 2-5A-dependent nuclease activity. Alternative possibilities to bring together the 2-5A binding site and the catalytic site within a single 37-kDa polypeptide have been proposed by Shetzline and Suhadolnik (14Whitaker J.R. Granum P.E. Anal. Biochem. 1980; 109: 156-159Crossref PubMed Scopus (491) Google Scholar). They imply non-conventional mechanisms such as ribosomal hopping or protein splicing, which have to be experimentally demonstrated. Whatever the case, RNase L proteolysis removes the Cys-rich region (in the protein kinase homology domain) of RNase L, which is believed to be involved in protein-protein interactions. It is, therefore, conceivable that the proteolysis of RNase L might alter its interaction with regulatory proteins or its compartmentalization. Such dysregulation could in turn lead to the degradation of cellular mRNA species, which are not normal targets of native RNase L. We thank Dr. J. Derencourt for help with the N-terminal sequencing, Prof. R. J. Suhadolnik (Temple University) for communicating unpublished data and for valuable discussions, and Dr. I. Robbins for editing the manuscript.
Referência(s)