Advertisement

Coordinated Induction of the Ubiquitin Conjugation Pathway Accompanies the Developmentally Programmed Death of Insect Skeletal Muscle (∗)

  • Arthur L. Haas
    Correspondence
    To whom correspondence should be addressed: Dept. of Biochemistry, Medical College of Wisconsin, 8701 Watertown Plank Rd., Milwaukee, WI 53226
    Affiliations
    From the Department of Biochemistry of the Medical College of Wisconsin, Milwaukee, Wisconsin 53226
    Search for articles by this author
  • Olga Baboshina
    Affiliations
    From the Department of Biochemistry of the Medical College of Wisconsin, Milwaukee, Wisconsin 53226
    Search for articles by this author
  • Bart Williams
    Footnotes
    Affiliations
    From the Department of Biochemistry of the Medical College of Wisconsin, Milwaukee, Wisconsin 53226
    Search for articles by this author
  • Lawrence M. Schwartz
    Affiliations
    Department of Biology, University of Massachusetts, Amherst, Massachusetts 01003
    Search for articles by this author
  • Author Footnotes
    ∗ This work was supported by U. S. Public Health Service Grants GM40458 and AG00492 (to L. M. S.) and GM34009 (to A. L. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
    Fellow of the Medical College of Wisconsin Summer Program for Undergraduate Research. Current address: Biology Dept., Massachusetts Institute of Technology, Cambridge, MA 02139.
Open AccessPublished:April 21, 1995DOI:https://doi.org/10.1074/jbc.270.16.9407
      The developmentally programmed cell death of abdominal intersegmental muscles in the tobacco hawkmoth Manduca sexta is coincident with a 10-fold induction of the polyubiquitin gene as a hormonally regulated event (Schwartz, L. M., Myer, A., Kosz, L., Engelstein, M., and Maier, C. (1990) Neuron 5, 411-419). Solid phase immunochemical assays measuring intersegmental muscle pools of free and conjugated ubiquitin reveal that the induction of polyubiquitin mRNA is accompanied by a proportional increase in total ubiquitin polypeptide. Ubiquitin conjugate pools increase 10-fold at eclosion, during which loss of muscle protein mass is maximum. A smaller but measurable increase in ubiquitin conjugates is observed earlier in pupal development coincident with a modest enhanced degradation of myofibrillar proteins. Accumulation of ubiquitin conjugates is accompanied by induction in the pathway for polypeptide ligation, including the activating enzyme (E1), several carrier protein (E2) isoforms, and ubiquitin:protein isopeptide ligase (E3). Both accumulation of ubiquitin polypeptide and the enzymes of the conjugation pathway are subject to regulation by declining titers of the insect molting hormone 20-hydroxyecdysone, which signals onset of programmed cell death in the intersegmental muscles. Thus, programmed cell death within the intersegmental muscles is accomplished in part by stimulation of the ubiquitin-mediated degradative pathway through a coordinated induction of ubiquitin and the enzymes responsible for its conjugation to yield proteolytic intermediates. This suggests enzymes required for ubiquitin conjugation may represent additional genes recruited for developmentally programmed death.

      INTRODUCTION

      The multi-enzyme pathway of ATP, ubiquitin-dependent proteolysis represents the principal cytosolic mechanism in eucaryotes for the specific degradation of short-lived and structurally abnormal proteins (
      • Hershko A.
      • Ciechanover A.
      ). Within this pathway, proteins are targeted for degradation through the ATP-coupled covalent ligation of the carboxyl terminus of ubiquitin (8.6 kDa) to ε-amino groups of lysine residues present on the protein substrate (discussed in Ref. 2). Enhanced exposure of lysine residues appears to be the underlying determinant for the specificity of ubiquitin ligation (
      • Dunten R.L.
      • Cohen R.E.
      • Gregori L.
      • Chau V.
      ,
      • Gregori L.
      • Marriott D.
      • Putkey J.A.
      • Means A.R.
      • Chau V.
      ). Conformational changes that increase the availability of potential sites of ubiquitin conjugation may arise by partial denaturation, errors in intracellular trafficking, or limited endoproteolytic cleavages that expose new destabilizing amino termini predisposed to recognition by the system (
      • Bachmair A.
      • Varshavsky A.
      ,
      • Johnson E.S.
      • Bartel B.
      • Seufert W.
      • Varshavsky A.
      ). In addition to this broad selectivity for abnormal proteins, ubiquitin-mediated degradation exhibits a marked specificity for a select subset of regulatory proteins for which covalent modification or alterations in ligand binding is thought to expose otherwise cryptic sites for conjugation (
      • Rechsteiner M.
      ). The latter class includes a variety of regulatory proteins such as cyclin B, the tumor supressor p53, c-Myc, c-Fos, and MAT 2α (
      • Ciechanover A.
      • DiGiuseppe J.A.
      • Bercovich B.
      • Orian A.
      • Richter J.D.
      • Schwartz A.L.
      • Brodeur G.M.
      ,
      • Glotzer M.
      • Murray A.W.
      • Kirschner M.W.
      ,
      • Band V.
      • De Caprio J.A.
      • Delmolino L.
      • Kulesa V.
      • Sager R.
      ,
      • Chen P.
      • Johnson P.
      • Sommer T.
      • Jentsch S.
      • Hochstrasser M.
      ). ATP, ubiquitin-dependent protein degradation has also been implicated in more global cellular processes such as skeletal muscle atrophy (
      • Wing S.S.
      • Goldberg A.L.
      ), spermatogenesis (
      • Kay G.F.
      • Ashworth A.
      • Penny G.D.
      • Dunlop M.
      • Swift S.
      • Brockdorff N.
      • Rastan S.
      ), and antigen processing in thymic stromal cells (
      • Goldberg A.L.
      • Rock K.L.
      ) for which current evidence suggests a general induction of the entire pathway.
      Programmed cell death offers some of the most dramatic examples of protein catabolism observed in animals; for example, loss of the entire tadpole tail during metamorphosis to the frog can occur in under 3 days (
      • Nieuwkoop P.D.
      • Faber J.
      ). However, at present little is known about the role of ATP, ubiquitin-dependent proteolysis during programmed cell death. It is reasonable to suspect that ubiquitin-mediated degradation is involved since dramatic increases in polyubiquitin message or protein have been observed in specific examples of programmed cell death (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ,
      • Lauzon R.J.
      • Patton C.W.
      • Weissman I.L.
      ,
      • Fahrbach S.
      • Schwartz L.M.
      ). One system that is amenable to a biochemical analysis of programmed cell death is the intersegmental muscle (ISM)1
      The abbreviations used are: ISM
      intersegmental muscle
      E1
      ubiquitin activating enzyme
      E2
      ubiquitin carrier protein (subscript denotes relative molecular weight by SDS-PAGE)
      E3
      ubiquitin-protein isopeptide ligase
      20-HE
      20-hydroxyecdysone
      PAGE
      polyacrylamide gel electrophoresis.
      of the tobacco hawkmoth Manduca sexta (
      • Lockshin R.A.
      ,
      • Schwartz L.M.
      • Kosz L.
      • Kay B.K.
      ). Groups of ISM span each of the abdominal segments of the larva and are composed of single syncytial cells that are each 5 mm long and up to 1 mm in diameter. Following pupation, the muscles in the first two and last two segments die. The ISM in the remaining four segments persist unmodified until day 15 of the normal 18 days of pupal/adult development. On day 15, the ISM begin to atrophy, resulting in a 40% loss of muscle mass without significant changes in the physiological properties of the cell (
      • Schwartz L.M.
      • Truman J.W.
      ,
      • Schwartz L.M.
      ). The ISMs are then required to perform the eclosion (emergence) behavior of the adult moth on day 18, after which the ISMs die during the subsequent 30 h (
      • Finlayson L.H.
      ,
      • Lockshin R.A.
      • Williams C.M.
      ).
      The trigger for both ISM atrophy and death is a decline in the circulating titer of the insect molting hormone 20-hydroxyecdysone (20-HE). An initial decline on day 14 of pupal/adult development initiates the atrophy program, while a further decline early on day 18 commits the muscles to die following adult eclosion later that day (
      • Schwartz L.M.
      • Truman J.W.
      ). If day 17 animals are given exogenous steroid, ISM death is delayed until the circulating titer of the hormone falls below the required threshold. Once the cells are committed to die, they can no longer be rescued by steroid administration. The ability of the ISM to die requires the repression and activation of a number of ecdysteroid-regulated genes (
      • Schwartz L.M.
      • Kosz L.
      • Kay B.K.
      ). If animals are treated with 20-HE on day 17, both death and the anticipated changes in gene expression are delayed. Two of the repressed genes encode actin and myosin heavy chain (
      • Schwartz L.M.
      • Jones M.E.E.
      • Kosz L.
      • Kuah K.
      ), while two up-regulated sequences encode apolipophorin III (
      • Sun D.
      • Ziegler R.
      • Milligan C.E.
      • Fahrbach S.
      • Schwartz L.M.
      ) and polyubiquitin (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ). In all species, the polyubiquitin genes encode multiple head-to-tail repeats of the ubiquitin coding sequence, which when translated results in an elevation of free ubiquitin monomers (
      • Jentsch S.
      • Seufert W.
      • Hauser H.P.
      ). The Manduca genome contains two polyubiquitin alleles that encode 9 and 18 repeats (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ).2
      A. Myer and L. M. Schwartz, manuscript in preparation.
      Translation of these polyubiquitin genes provides the cell with a powerful amplification system for increasing the molar concentration of ubiquitin within the cell.
      The current studies were initiated to exploit existing immunochemical methods for the quantitation of free and conjugated ubiquitin pools (
      • Haas A.L.
      • Bright P.M.
      ,
      • Haas A.L.
      • Bright P.M.
      ) within intersegmental muscles during the final stages of metamorphosis. The results indicate that total ubiquitin levels, predominantly in the form of conjugates, increase exponentially at the time of cell death in response to a coordinated induction of both polyubiquitin and the enzymes required for conjugate formation. This developmental response is regulated by the same hormonal trigger required for initiation of cell death. The coordinated induction of the conjugation pathway supports a role for ubiquitin-mediated degradation in developmentally programmed death and suggests that other putative cell death genes may represent enzyme(s) of the ubiquitin conjugation pathway.

      MATERIALS AND METHODS

      Bovine ubiquitin and protein A (Sigma) were radioiodinated by the chloramine-T procedure (
      • Haas A.L.
      • Warms J.V.
      • Hershko A.
      • Rose I.A.
      ) using carrier-free Na125I obtained from Amersham Radiochemicals. Homogeneous ubiquitin-conjugating enzymes were isolated by combined affinity and fast protein liquid chromatography methods (
      • Haas A.L.
      • Bright P.M.
      ) from rabbit reticulocytes prepared by phenylhydrazine induction (
      • Haas A.L.
      • Rose I.A.
      ). Affinity-purified rabbit polyclonal antibodies specific for conjugated ubiquitin were those previously described (
      • Haas A.L.
      • Bright P.M.
      ,
      • Haas A.L.
      • Bright P.M.
      ,
      • Haas A.L.
      • Murphy K.E.
      • Bright P.M.
      ). The insect molting hormone 20-hydroxyecdysone was purchased from Sigma. Protein was determined by the method of Lowry et al. (
      • Lowry O.H.
      • Rosebrough N.J.
      • Farr A.L.
      • Randall R.J.
      ), using bovine serum albumin as standard.

      Preparation of Intersegmental Muscle Samples

      Intersegmental muscles were rapidly dissected in ice-cold saline from animals at various stages of pupal-adult development before and after the onset of cell death (
      • Schwartz L.M.
      • Truman J.W.
      ). Excised muscles were flash frozen in liquid nitrogen and stored at −80°C until analyzed to prevent degradation of the conjugate pool and to preserve endogenous conjugating activities. After weighing the samples, extracts were prepared from individual muscles by mincing the frozen tissue on a glass plate chilled to −60°C. The minced tissue was disrupted by sonication for 40 s in 4 volumes of ice-cold homogenizing buffer composed of 50 m M Tris-Cl (pH 7.5), 0.25 M sucrose, 1% (w/v) SDS, 1 m M tosyl- L-lysine chloromethyl ketone, 1 m ML-tosylamido-2-phenylethyl chloromethyl ketone, 5 m M EDTA, and 5 m M N-ethylmaleimide (
      • Riley D.A.
      • Bain J.L.
      • Ellis S.
      • Haas A.L.
      ). Crude extracts were immediately boiled for 10 min and then centrifuged for 5 min at 14,000 × g to remove insoluble debris. This method of sample preparation has been previously shown to quantitatively preserve conjugates in proteolytically active rabbit reticulocyte (
      • Haas A.L.
      • Bright P.M.
      ) and rat skeletal muscle extracts (
      • Riley D.A.
      • Bain J.L.
      • Ellis S.
      • Haas A.L.
      ) and was independently confirmed for the present samples. In studies requiring activity assays for endogenous conjugating enzymes, extracts were prepared from contralateral muscles using homogenizing buffer supplemented with 1 m M dithiothreitol from which SDS, EDTA, and the protease inhibitors had been omitted.

      Quantitation of Free and Conjugated Ubiquitin Pools

      The absolute content of free ubiquitin within the intersegmental muscle extracts was determined by Western blotting using affinity-purified rabbit polyclonal antibodies against ubiquitin (
      • Haas A.L.
      • Bright P.M.
      ). Conjugate levels were determined by a solid phase dot blot assay against standards prepared as previously described (
      • Haas A.L.
      • Bright P.M.
      ). Samples were diluted with 50 m M Tris-Cl (pH 7.5) containing 0.15 M NaCl and 0.01% (w/v) SDS to be within the empirically determined linear response range of the assay. Immunospecifically bound antibody was detected by autoradiography following incubation with 125I-protein A (
      • Haas A.L.
      • Bright P.M.
      ). The resulting dot blot autoradiograms were quantitated densitometrically using a BioTek EL308 microplate reader fitted with a 550-nm filter. The latter refinement significantly improved the sensitivity and reproducibility of the assay over previous methods (
      • Haas A.L.
      • Bright P.M.
      ).

      125I-Ubiquitin Assays for Conjugating Enzymes

      Endogenous intersegmental muscle E1 and E2 isozymes were assayed by their ability to form 125I-ubiquitin thiol ester (
      • Haas A.L.
      • Bright P.M.
      ,
      • Haas A.L.
      • Warms J.V.
      • Hershko A.
      • Rose I.A.
      ). Briefly, extracts were incubated 1 min at 37°C in a 50-μl final volume containing 50 m M Tris-Cl (pH 7.5), 1 m M ATP, 10 m M MgCl2, 1 m M dithiothreitol, and 5 μM125I-ubiquitin (approximately 10,000 cpm/pmol). Stoichiometric formation of E1 and E2 125I-ubiquitin thiol esters were quantitated after resolving the incubations by non-reducing SDS-PAGE in the cold, cutting the resulting bands from the dried gel and determining the associated 125I-ubiquitin by γ counting (
      • Haas A.L.
      • Bright P.M.
      ,
      • Haas A.L.
      • Warms J.V.
      • Hershko A.
      • Rose I.A.
      ). Ubiquitin-protein ligase (E3) activity was determined in incubations similar to those for thiol ester formation with the exception that reactions additionally contained 10 m M creatine phosphate and 10 IU/ml rabbit muscle creatinine phosphokinase as an ATP regenerating system (
      • Haas A.L.
      • Rose I.A.
      ). Initial rates of 125I-ubiquitin conjugation to endogenous intersegmental muscle proteins were measured by resolving the samples on SDS-PAGE, cutting the lanes from the gel, and quantitating the associated radioactivity by γ counting. The concentration of 125I-ubiquitin present in both the thiol ester and conjugation assays was sufficient to obviate isotope dilution by endogenous unlabeled ubiquitin in the extracts, based on prior pool measurements.

      RESULTS

      Ubiquitin Pools Increase Following Eclosion

      Two distinct phases were consistently observed when free and conjugated ubiquitin pools were quantitated in samples obtained at the indicated times during the late pupal stage and following eclosion (Fig. 1). In the first phase, total ubiquitin remained constant from late in day 14 through early in day 18 (closed circles); however, there was a statistically significant 2-fold increase in the absolute size of the conjugate pool over the same time frame, resulting from redistribution of total polypeptide from the free to the conjugated form. This shift between pools was reflected in the increased fraction of total ubiquitin conjugated, Fig. 1 (open circles), and coincided with the initial increase in specific myofibrillar protein degradation occurring during this time (
      • Schwartz L.M.
      • Truman J.W.
      ,
      • Schwartz L.M.
      ). Western blot analysis of the molecular weight distribution of ubiquitin adducts indicated a general accumulation of the total population of adducts present late in day 14 (not shown). Within the limits of resolution for the Western blots, this indicates that the same population of protein targets is subject to enhanced conjugation.
      Figure thumbnail gr1
      Figure 1:Intersegmental muscle ubiquitin pools during late development. Free and conjugated ubiquitin pools within intersegmental muscle extracts were quantitated as described under “Materials and Methods” from which total ubiquitin (solid circles) and the percent conjugated (open circles) were calculated. Each point represents the mean ± S.E. of pooled muscles from five animals. Time of adult eclosion is noted in the figure.
      The second phase of the response was characterized by a 10-fold increase in the total ubiquitin pool from late in day 18 until 17 h post-eclosion, at which time tissue sampling became difficult. This increase in total ubiquitin agrees temporally and quantitatively with the 10-fold increase in polyubiquitin mRNA previously reported (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ). During the second phase of the response, the fractional level of conjugation exhibits an abrupt elevation to a final value of 93% by 17 h post-eclosion (Fig. 1). In contrast, free ubiquitin levels remained relatively constant at 35 ± 2 pmol/mg protein throughout both phases of the time course shown in Fig. 1. Therefore, newly synthesized ubiquitin arising from induction of the polyubiquitin gene partitions directly into the conjugate pool, suggesting that the normal homeostatic mechanism for maintaining the balance between free and conjugated ubiquitin is disabled (
      • Haas A.L.
      • Bright P.M.
      ). Western blots of the samples obtained during the second phase again revealed no change in the distribution of conjugates (not shown). Qualitatively similar results were observed in parallel experiments, although the exact timing of events varied due to uncertainty in the initial staging of pupa.

      Coordinated Induction of Conjugating Activity Precedes Eclosion

      Extracts were prepared from muscles contralateral to those used for the ubiquitin pool assays in Fig. 1 to test whether induction of ubiquitin conjugating activity coincided with the increase in ubiquitin adducts (Fig. 2). For each time point, equal volumes of muscle extract were pooled and then dialyzed (12-kDa exclusion) against 50 m M Tris-Cl (pH 7.5) containing 1 m M dithiothreitol to remove metabolites and endogenous free ubiquitin. Separate control studies indicated that reticulocyte E1 and E214K, when added to parallel pooled extracts, were stable during dialysis so that loss of conjugating activity by inactivation could be precluded (not shown). Initial rates of conjugation by the dialyzed extracts were measured in the presence of saturating 125I-ubiquitin (5 μM) to obviate isotope dilution by any remaining endogenous ubiquitin not removed during dialysis. The resulting free and conjugated 125I-ubiquitin in the incubations were resolved by SDS-PAGE and then quantitated by cutting lanes from the dried gel and determining associated radioactivity by γ counting (
      • Haas A.L.
      • Rose I.A.
      ). Replicate determinations for rates of conjugation for a given pooled sample were generally within 5% of the mean value.
      A 1.5-fold increase in conjugating activity is observed during day 15 (Fig. 2), which coincides with the elevation in fractional level of conjugation observed during the initial phase of the response (Fig. 1), after which conjugating activity remained relatively constant until day 18. A significant increase in conjugating activity occurred early in day 18 and continued through eclosion just prior to the beginning of day 19. The latter stimulation in conjugating activity preceded the increase in accumulation of ubiquitin conjugates occurring during the second phase of the response (Fig. 1). The total elevation in conjugating activity from late day 14 to the maximum at eclosion was 5-fold (Fig. 2), in good agreement with the overall 10-fold increase in the conjugate pool from Fig. 1, suggesting that the net accumulation of conjugates is kinetically driven. Qualitatively similar temporal results were observed in two other replicate experiments (not shown), although the absolute timing of eclosion varied due to some uncertainty in the initial staging of pupa. The small decline in conjugation rate at 4 h post-eclosion shown in Fig. 2 was not consistently observed; indeed, two other experiments showed increased activity at 4 h post-eclosion (not shown). In the latter two experiments, the absolute pools of ubiquitin conjugates were smaller than that shown in Fig. 1, suggesting that the enzymes required for ubiquitin conjugation may also be subject to autodegradation by the system in very actively atrophying muscles, resulting in a net decline in rate of ubiquitin conjugation following eclosion. Despite this discrepancy, there was a rough correlation among the three experiments between rate of conjugation and size of the conjugate pool following eclosion.
      Figure thumbnail gr2
      Figure 2:Changes in 125I-ubiquitin conjugation activity during late development. Initial rates of conjugation were determined in pooled extracts prepared from contralateral intersegmental muscles from those used in the study shown in Fig. 1. Assays of 20 min at 37°C were performed as described under “Materials and Methods” using 360 μg of muscle extract protein.
      The assay results of Fig. 2 reflected the ubiquitin-protein ligase (E3)-dependent step in the cell-free extracts since rates of conjugation were unaffected by addition of homogeneous rabbit reticulocyte E1 and E214Kor by increasing the concentration of 125I-ubiquitin (not shown). Independent complimentation experiments confirmed that reticulocyte E1 and E214Kfully supported the intersegmental muscle conjugating activity. Since the assays in Fig. 2 necessarily utilize endogenous extract protein as substrates for ligation, the data do not distinguish between increases in net E3 activity and increases in the pool of proteins susceptible to ubiquitin ligation. However, since the pattern of ubiquitin adducts revealed by immunoblots was not qualitatively different during the time of enhanced conjugating activity, it is unlikely the observed increase in rate is attributable to appearance of a distinct pool of target protein substrates. The results of Fig. 2 do demonstrate that conjugating activity increased prior to the second phase induction of total ubiquitin at the time of eclosion and that a small but consistent increase in conjugating activity was observed between days 15 and 16 (Fig. 2) that correlated with the initial phase increase in fractional level of conjugation (Fig. 1).

      Coordinated Induction of E1 and E2 Isozymes Parallels Increased Rates of Conjugation

      Endogenous levels of E1 and E2 isozymes present in the pooled extracts were qualitatively estimated autoradiographically by their stoichiometric formation of 125I-ubiquitin thiol esters in brief incubations (Ref. 2 and Fig. 3). This functional thiol ester assay for E1 and E2 in crude extracts exploits the rapid formation of these intermediates compared to the relatively slower rate of subsequent conjugation (
      • Haas A.L.
      • Bright P.M.
      ). Thiol esters can be distinguished from conjugates by the former's sensitivity to thiolytic cleavage by heating in the presence of β-mercaptoethanol (
      • Haas A.L.
      • Bright P.M.
      ). Most of the bands present in Fig. 3 are thiol esters since they were absent from parallel control gels run under reducing conditions (not shown). The dense band at the top of each lane represents rapidly formed high molecular weight conjugates resistant to boiling in the presence of β-mercaptoethanol. Finally, control studies confirmed that the dialyzed extracts contained no inhibitors of reticulocyte E1 or E214Kthiol ester formation (not shown).
      Figure thumbnail gr3
      Figure 3:Functional assay of E1 and E2 isozymes during late development. Endogenous E1 and E2 isozymes present in muscle extracts were quantitated by their stoichiometric formation of 125I-ubiquitin thiol esters in incubations identical to those of Fig. 2 with the exceptions that reactions were for 1 min and the samples were resolved in the cold by non-reducing SDS-PAGE (see “Materials and Methods”). To the right of the figure are shown migration positions for ubiquitin, E1, and selected rabbit reticulocyte E2 isozymes run in parallel.
      Manduca E1 thiol ester comigrates with the corresponding reticulocyte activating enzyme intermediate, indicating that the insect enzyme is of similar molecular weight to that of vertebrates. Intersegmental muscle extracts contained a range of lower molecular weight thiol ester adducts that probably represent E2 isozymes since several comigrated with reticulocyte E2 isozyme standards (Fig. 3). Intersegmental muscle extracts contained three major E2 thiol ester bands that migrated with or slightly below rabbit E214K, the principal isozyme present in reticulocytes (
      • Haas A.L.
      • Bright P.M.
      ,
      • Pickart C.M.
      • Rose I.A.
      ). Depending on running conditions, thiol ester gels can reveal spurious additional bands arising from unfolding artifacts peculiar to the non-reducing SDS-PAGE system (
      • Haas A.L.
      • Bright P.M.
      ,
      • Haas A.L.
      • Bright P.M.
      • Jackson V.E.
      ). However, the two thiol ester bands migrating just below the reticulocyte E214Kstandard are probably not running artifacts since their autoradiographic intensities are not proportional to that comigrating with the reticulocyte marker (Fig. 3). Therefore, these two putative isoforms may represent cell death-specific E2 species unique to ISM.
      Thiol ester bands present in the day 14.8 pooled samples were just at the limit of detection for the exposure of the original autoradiogram from which Fig. 3 was derived. Small increases in thiol esters formed with E1, E214K, and the two smaller putative isoforms occurred during day 15 (Fig. 3). Induction of these bands coincided with the increases in conjugating activity (Fig. 2) and percent conjugation (Fig. 1) observed during the initial phase stimulation of myofibrillar protein degradation. Levels of these four thiol ester bands remained relatively constant until late in day 17 when a second wave of induction occurred that continued through eclosion. Additional bands of thiol esters are also induced during this second stage, one pair of which comigrate with rabbit E220K(Fig. 3). The second stage of induction coincided with the larger second phase increase in conjugating activity occurring over this same time frame (Fig. 2). The drop in conjugating activity observed 4 h post-eclosion (Fig. 2) parallels a reduction in the thiol ester bands in Fig. 3. Therefore, the results in Fig. 3 demonstrate that E1 and E2 isozymes coordinately increase during the same time frames in which overall conjugating activity and fractional level of conjugation are also elevated.

      Induction of Ubiquitin Conjugates Is Under Hormonal Regulation by 20-Hydroxyecdysone

      As reviewed in the Introduction, the onset for programmed cell death within intersegmental muscles is triggered by a natural decline in the circulating titer of 20-HE occurring late on day 17 that induces transcription of a subset of putative “cell death genes” (
      • Schwartz L.M.
      ). Cell death and induction of these genes, including that for polyubiquitin, can be delayed by injection of exogenous hormone prior to the programmed decline in its circulating titer (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ,
      • Schwartz L.M.
      • Truman J.W.
      ). The data of Table I demonstrate that artificial manipulation of 20-HE also blocks both the accumulation of ubiquitin conjugates and their rates of formation. Pupa were injected with either 20-HE or an equal volume of carrier 10% ethanol early in day 17, prior to the decline in the natural titer of the molt hormone. At the indicated times, intersegmental muscles were harvested, and the ubiquitin pools were quantitated as in Fig. 1. Contralateral muscles were assayed for 125I-ubiquitin conjugating activity as described in the legend of Fig. 2.
      Table I:Effect of 20-hydroxyecdysone on intersegmental muscle ubiquitin pools and rates of conjugation
      Intersegmental muscles at 4 h post-eclosion showed a significant increase in both total ubiquitin and the fractional level of conjugation compared to reference muscles harvested late on day 17 (Table I), in agreement with the results of Fig. 1. In contrast, animals injected early on day 17 with 20-HE but assayed 4 h after the normal time of eclosion contained a total ubiquitin pool and percent conjugation similar to that of the day 17 control muscles. Both the induction of total ubiquitin and the increase in percent conjugation must be under the same hormonal regulation that signals programmed cell death in this tissue since control muscles injected with carrier exhibited ubiquitin pools and percent conjugation statistically identical to untreated 4 h post-eclosion samples. The increase in rate of in vitro E3-dependent conjugation that preceded eclosion (Fig. 2) was also blocked by injection of 20-HE but not by treatment with carrier alone (Table I). Parallel experiments similar to that of Fig. 3 showed that 20-HE, but not carrier, also blocked the marked increase in E1 and E2 isozyme thiol esters observed between late day 17 and eclosion (not shown). The data in indicate that the hormonal regulation for onset of cell death extends to a coordinated induction of ubiquitin and the enzyme(s) responsible for its ligation to intracellular protein targets.

      DISCUSSION

      Necrosis within eucaryotes is the passive lethal response to cellular injury resulting from thermal or chemical stress, ischemia, and metabolic inhibitors. This process usually involves disruption of membrane integrity followed by ionic disequilibrium, cellular swelling, and ultimately lysis. In contrast, the majority of cell deaths observed within animals involves an active metabolic process that is dependent on de novo gene expression (
      • Schwartz L.M.
      • Osborne B.A.
      ). In most cases, this process occurs with the morphology of apoptosis, which includes cellular condensation, genomic DNA degradation, cell surface blebbing to form membrane-bound apoptotic bodies, and ultimately phagocytosis (
      • Cohen J.J.
      ). Death of the intersegmental muscles, as well as many other cells including mammalian neurons, also depends on gene expression but proceeds with a morphology that is distinct from apoptosis (
      • Schwartz L.M.
      • Jones M.E.E.
      • Kosz L.
      • Kuah K.
      ). In Manduca, this form of programmed cell death involves dramatic elevations in the expression of the polyubiquitin gene (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ).
      The present study represents the most detailed examination to date of the potential role of ubiquitin-mediated degradation in cellular remodeling. We demonstrate that induction of polyubiquitin mRNA following eclosion during the programmed cell death of ISM is accompanied by a proportional accumulation of the mature ubiquitin polypeptide (Fig. 1 and). This new, expanded pool of ubiquitin partitions directly into cellular conjugates since free polypeptide remains constant throughout the late pupal stage and eclosion (Fig. 1). We have previously shown that cultured cells and tissues regulate the ratio between free to conjugated ubiquitin through poorly understood mechanisms of homeostasis, except under conditions of stress or other factors correlated with enhanced degradation (
      • Haas A.L.
      ). The sharp increase in the fractional level of ubiquitin conjugates in ISM during terminal development indicates that comparable regulation is either absent or severely attenuated. Kinetic models based on experimental observations require that degradative flux through the ubiquitin pathway be proportional to the concentration of ubiquitin conjugates (
      • Haas A.L.
      ). Accordingly, the marked increase in these degradative intermediates immediately prior to eclosion coincides with the enhanced rate of protein loss within ISM. Correlation between the more modest increase in conjugates during day 15 and an elevated rate of specific myofibrillar protein degradation (
      • Schwartz L.M.
      • Truman J.W.
      ,
      • Schwartz L.M.
      ) suggests that the ubiquitin-mediated pathway is also responsible for this developmentally programmed atrophy, consistent with its role in the proteolysis of myofibrillar proteins in rat skeletal muscles (
      • Wing S.S.
      • Goldberg A.L.
      ,
      • Haas A.L.
      • Riley D.A.
      ).3
      Wing, S., Haas, A. L., and Goldberg, A. L. (1995) Biochem. J., in press.
      A previously unreported increase in net ubiquitin conjugating activity within ISM precedes the marked accumulation of conjugates on day 19 (Fig. 2). Enhanced ubiquitin conjugation during day 18 is attributable to induction of isopeptide ligase (E3) activity under assay conditions for which neither E1 nor E2 was rate limiting. Stoichiometric thiol ester assays indicate that increased E3 activity is coincident with induction of E1 and several E2 isozymes, two of which are novel isoforms of molecular weight less than that of the major E214Kpolypeptide (Fig. 3). However, the enhanced capacity to ligate ubiquitin does not result in a significant accumulation of conjugates until induction of ubiquitin on day 19. The concerted induction of E1 and E2 is probably required to ensure substrate specificity of ubiquitin ligation since induction of E3 alone could potentially result in shifting the rate-limiting step of conjugation from E3 to an earlier reaction in the ligation pathway. Because substrate specificity resides in the E3-catalyzed step, kinetic specificity for target protein conjugation would otherwise be lost.
      Like the induction of polyubiquitin message (
      • Schwartz L.M.
      • Myer A.
      • Kosz L.
      • Engelstein M.
      • Maier C.
      ) and the accumulation of ubiquitin protein, the increase in E3 measured as net ubiquitin conjugating activity as well as E1 and E2 levels measured by thiol ester formation are also regulated by decline in the titer of 20-HE (and accompanying text). During this time, an increase in multicatalytic protease activity, responsible for the ATP-dependent degradation of ubiquitin conjugates, and the appearance of four novel multicatalytic protease subunits is also observed.4,5
      Jones, M. E. E., Haire, M. F., Kloetzel, P.-M., Mykles, D., and Schwartz, L. M. Dev. Biol., in press.
      Haire, M. F., Clark, J. J., Jones, M. E. E., Hendil, K. B., Schwartz, L. M., and Mykles, D. L. (1995) Arch. Biochem. Biophys., in press.
      Therefore, commitment to ISM cell death represents a coordinated expansion in the capacity of the cell to carry out ubiquitin-mediated degradation. This suggests that other putative cell death genes probably will be found to be additional components of this degradative pathway, particularly that required for ubiquitin conjugation.
      An increase in the expression of polyubiquitin message or ubiquitin protein at the time of programmed cell death is observed in a number of tissues, including Manduca neurons (
      • Fahrbach S.
      • Schwartz L.M.
      ) and tunicates (
      • Lauzon R.J.
      • Patton C.W.
      • Weissman I.L.
      ). Additionally, Delic et al. (
      • Delic J.
      • Morange M.
      • Magdelenat H.
      ) have demonstrated that antisense inhibition of polyubiquitin gene expression blocks the γ irradiation-induced death of human peripheral lymphocytes (
      • Delic J.
      • Morange M.
      • Magdelenat H.
      ). However, many cells can undergo programmed cell death without detectable increases in ubiquitin expression, including Drosophila ommatidia (
      • Wolff T.
      • Ready D.F.
      ), rat sympathetic neurons (
      • Martin D.P.
      • Ito A.
      • Horigome K.
      • Lampe P.A.
      • Johnson Jr., E.M.
      ), PC12 cells (
      • D'Mello S.R.
      • Galli C.
      ), and mouse T cell hybridomas (
      • Schwartz L.M.
      • Smith S.W.
      • Jones M.E.E.
      • Osborne B.A.
      ). In these tissues, basal levels of ubiquitin may be sufficient to meet the proteolytic demands of cell death. Alternatively, since cells undergoing apoptosis are rapidly phagocytosed (
      • Cohen J.J.
      ), much of the degradative responsibility may reside with phagocytic mechanisms rather than on the dying cell in these latter cases; consequently, induction of the ubiquitin-mediated pathway may not be a general feature of all examples of programmed cell death. It should also be noted that some tissues constitutively express very high levels of ubiquitin-protein conjugates without a commitment to die, such as certain neurosecretory cells in Manduca (
      • Fahrbach S.
      • Schwartz L.M.
      ) and rat (
      • Schueler P.A.
      • Elde R.P.
      ). In these cases, elevation in ubiquitin-dependent proteolysis may be required for lineage-specific demands such as neuropeptide processing.
      Elevations in ubiquitin expression are not only seen with programmed cell death but also with certain pathological cell deaths (
      • Mori H.
      • Kondo J.
      • Ihara Y.
      ,
      • Mayer R.J.
      • Arnold J.
      • Laszlo L.
      • Landon M.
      • Lowe J.
      ). When cerebral neurons are subjected to transient ischemia, massive cell death occurs with reperfusion. Mortality is accompanied by a dramatic increase in ubiquitin expression in the surviving neurons (
      • Hayashi T.
      • Takada K.
      • Matsuda M.
      ,
      • Kato H.
      • Chen T.
      • Liu X.H.
      • Nakata N.
      • Kogure K.
      ,
      • Yee W.M.
      • Frim D.M.
      • Isacson O.
      ). Cells capable of mobilizing the ubiquitin-dependent proteolytic pathway may be better able to deal with the generation of misfolded or otherwise abnormal proteins following oxidative stress. In addition, ubiquitin is a major component of the neurofibrillary tangles characteristic of Alzheimer's disease (
      • Mori H.
      • Kondo J.
      • Ihara Y.
      ,
      • Tabaton M.
      • Cammarata S.
      • Mancardi G.
      • Manetto V.
      • Autilio-Gambetti L.
      • Perry G.
      • Gambetti P.
      ). Other neurodegenerative diseases, such as amyotrophic lateral sclerosis (
      • Wightman G.
      • Anderson V.E.
      • Martin J.
      • Swash M.
      • Anderton B.H.
      • Neary D.
      • Mann D.
      • Luthert P.
      • Leigh P.N.
      ), motor neuron disease (
      • Lowe J.
      • Blanchard A.
      • Morrell K.
      • Lennox G.
      • Reynolds L.
      • Billett M.
      • Landon M.
      • Mayer R.J.
      ), Parkinsons disease (
      • Lowe J.
      • Blanchard A.
      • Morrell K.
      • Lennox G.
      • Reynolds L.
      • Billett M.
      • Landon M.
      • Mayer R.J.
      ), and dementia pugilistica (
      • Pickart C.M.
      • Graziani L.A.
      • Dosch S.F.
      ) also accumulate ubiquitin protein within cerebral structures. These data argue that ubiquitin may serve a role in both the normal and pathological death of cells, most notably neurons. Since it is not possible to assess changes in the levels and activities of the ATP, ubiquitin-dependent proteolytic pathway in individuals with neurodegenerative disorders, the use of animal models may facilitate the development of testable hypotheses that can be examined in a clinical setting.
      Ubiquitin-mediated proteolysis has previously been associated with the degradation of specific proteins or a limited subpopulation of targets. Such specific enhanced degradation results from changes in the susceptibility ot the protein target to ubiquitination and subsequent degradation without observable alterations in the capacity of the cell to catalyze this pathway. The present results demonstrate ubiquitin-mediated degradation can also be recruited for the global atrophy of a tissue such as the Manduca intersegmental muscle during programmed cell death. Under the latter conditions, general enhanced degradation arises by the concerted induction of a constellation of genes coding for ubiquitin, the enzymes responsible for conjugation of the polypeptide to target proteins, and subunits of the 26 S multicatalytic proteasome required for degradation of the resulting ubiquitin conjugates. Thus, the general enhanced degradation typified by erythroid maturation (
      • Pickart C.M.
      • Graziani L.A.
      • Dosch S.F.
      ), embryogenesis (
      • Pickart C.M.
      • Summers R.G.
      • Shim H.
      • Kasperek E.M.
      ), and programmed cell death (this study) requires the coordinated induction of enzyme activities representing the complete degradative pathway for the respective developmental transitions. This conclusion expands the potential roles for the ubiquitin pathway in cellular regulation.

      Acknowledgments

      We are indebted to Patricia Reback for technical help in the experiments reported here and to Dr. John Buckner for provision of Manduca eggs.

      REFERENCES

        • Hershko A.
        • Ciechanover A.
        Annu. Rev. Biochem. 1992; 61: 761-807
        • Haas A.L.
        • Bright P.M.
        J. Biol. Chem. 1988; 263: 13258-13267
        • Dunten R.L.
        • Cohen R.E.
        • Gregori L.
        • Chau V.
        J. Biol. Chem. 1991; 266: 3260-3267
        • Gregori L.
        • Marriott D.
        • Putkey J.A.
        • Means A.R.
        • Chau V.
        J. Biol. Chem. 1987; 262: 2562-2567
        • Bachmair A.
        • Varshavsky A.
        Cell. 1989; 56: 1019-1032
        • Johnson E.S.
        • Bartel B.
        • Seufert W.
        • Varshavsky A.
        EMBO J. 1992; 11: 497-505
        • Rechsteiner M.
        Cell. 1991; 66: 615-618
        • Ciechanover A.
        • DiGiuseppe J.A.
        • Bercovich B.
        • Orian A.
        • Richter J.D.
        • Schwartz A.L.
        • Brodeur G.M.
        Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 139-143
        • Glotzer M.
        • Murray A.W.
        • Kirschner M.W.
        Nature. 1991; 349: 132-138
        • Band V.
        • De Caprio J.A.
        • Delmolino L.
        • Kulesa V.
        • Sager R.
        J. Virol. 1991; 65: 6671-6676
        • Chen P.
        • Johnson P.
        • Sommer T.
        • Jentsch S.
        • Hochstrasser M.
        Cell. 1993; 74: 357-369
        • Wing S.S.
        • Goldberg A.L.
        Am. J. Physiol. 1993; 264: E668-E676
        • Kay G.F.
        • Ashworth A.
        • Penny G.D.
        • Dunlop M.
        • Swift S.
        • Brockdorff N.
        • Rastan S.
        Nature. 1991; 354: 486-489
        • Goldberg A.L.
        • Rock K.L.
        Nature. 1992; 357: 375-379
        • Nieuwkoop P.D.
        • Faber J.
        Normal Table of Xenopus Laevis (Daudin). North Holland Publishing Co., Amsterdam1967
        • Schwartz L.M.
        • Myer A.
        • Kosz L.
        • Engelstein M.
        • Maier C.
        Neuron. 1990; 5: 411-419
        • Lauzon R.J.
        • Patton C.W.
        • Weissman I.L.
        Cell Tissue Res. 1993; 272: 115-127
        • Fahrbach S.
        • Schwartz L.M.
        J. Comp. Neurol. 1994; 343: 464-482
        • Lockshin R.A.
        J. Insect Physiol. 1969; 15: 1505-1516
        • Schwartz L.M.
        • Kosz L.
        • Kay B.K.
        Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 6594-6598
        • Schwartz L.M.
        • Truman J.W.
        Dev. Biol. 1983; 99: 103-114
        • Schwartz L.M.
        J. Neurobiol. 1982; 23: 1312-1326
        • Finlayson L.H.
        Q. J. Microsc. Sci. 1956; 97: 215-233
        • Lockshin R.A.
        • Williams C.M.
        J. Insect Physiol. 1965; 11: 123-133
        • Schwartz L.M.
        • Jones M.E.E.
        • Kosz L.
        • Kuah K.
        Dev. Biol. 1993; 158: 448-455
        • Sun D.
        • Ziegler R.
        • Milligan C.E.
        • Fahrbach S.
        • Schwartz L.M.
        J. Neurobiol. 1995; 26: 119-129
        • Jentsch S.
        • Seufert W.
        • Hauser H.P.
        Biochim. Biophys. Acta. 1991; 1089: 127-139
        • Haas A.L.
        • Bright P.M.
        J. Biol. Chem. 1985; 260: 12464-12473
        • Haas A.L.
        • Bright P.M.
        J. Biol. Chem. 1987; 262: 345-351
        • Haas A.L.
        • Warms J.V.
        • Hershko A.
        • Rose I.A.
        J. Biol. Chem. 1982; 257: 2543-2548
        • Haas A.L.
        • Rose I.A.
        Proc. Natl. Acad. Sci. U. S. A. 1981; 78: 6845-6848
        • Haas A.L.
        • Murphy K.E.
        • Bright P.M.
        J. Biol. Chem. 1985; 260: 4694-4703
        • Lowry O.H.
        • Rosebrough N.J.
        • Farr A.L.
        • Randall R.J.
        J. Biol. Chem. 1951; 193: 265-275
        • Riley D.A.
        • Bain J.L.
        • Ellis S.
        • Haas A.L.
        J. Histochem. Cytochem. 1988; 36: 621-632
        • Pickart C.M.
        • Rose I.A.
        J. Biol. Chem. 1985; 260: 1573-1581
        • Haas A.L.
        • Bright P.M.
        • Jackson V.E.
        J. Biol. Chem. 1988; 263: 13268-13275
        • Schwartz L.M.
        • Osborne B.A.
        Immunol. Today. 1993; 14: 582-590
        • Schwartz L.M.
        • Smith S.W.
        • Jones M.E.E.
        • Osborne B.A.
        Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 980-984
        • Haas A.L.
        Rechsteiner M. Ubiquitin. Plenum Press, New York1988: 173-206
        • Haas A.L.
        • Riley D.A.
        Schlesinger M. Hershko A. Current Communications in Molecular Biology. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1988: 178-185
        • Delic J.
        • Morange M.
        • Magdelenat H.
        Mol. Cell. Biol. 1993; 13: 4875-4883
        • Wolff T.
        • Ready D.F.
        Development. 1991; 113: 825-839
        • Martin D.P.
        • Ito A.
        • Horigome K.
        • Lampe P.A.
        • Johnson Jr., E.M.
        J. Neurobiol. 1992; 23: 1205-1220
        • D'Mello S.R.
        • Galli C.
        Neuroreport. 1993; 4: 355-358
        • Cohen J.J.
        Immunol. Today. 1993; 14: 126-130
        • Schueler P.A.
        • Elde R.P.
        J. Chem. Neuroanat. 1988; 1: 165-173
        • Mori H.
        • Kondo J.
        • Ihara Y.
        Science. 1987; 235: 1641-1644
        • Mayer R.J.
        • Arnold J.
        • Laszlo L.
        • Landon M.
        • Lowe J.
        Biochim. Biophys. Acta. 1991; 1089: 141-157
        • Hayashi T.
        • Takada K.
        • Matsuda M.
        J. Neurosci. Res. 1992; 31: 561-564
        • Kato H.
        • Chen T.
        • Liu X.H.
        • Nakata N.
        • Kogure K.
        Brain Res. 1993; 619: 339-343
        • Yee W.M.
        • Frim D.M.
        • Isacson O.
        Exp. Brain Res. 1993; 94: 193-202
        • Tabaton M.
        • Cammarata S.
        • Mancardi G.
        • Manetto V.
        • Autilio-Gambetti L.
        • Perry G.
        • Gambetti P.
        Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 2098-2102
        • Wightman G.
        • Anderson V.E.
        • Martin J.
        • Swash M.
        • Anderton B.H.
        • Neary D.
        • Mann D.
        • Luthert P.
        • Leigh P.N.
        Neurosci. Lett. 1992; 139: 269-274
        • Lowe J.
        • Blanchard A.
        • Morrell K.
        • Lennox G.
        • Reynolds L.
        • Billett M.
        • Landon M.
        • Mayer R.J.
        J. Pathol. 1988; 155: 9-15
        • Pickart C.M.
        • Graziani L.A.
        • Dosch S.F.
        Arch. Biochem. Biophys. 1989; 272: 114-121
        • Pickart C.M.
        • Summers R.G.
        • Shim H.
        • Kasperek E.M.
        Dev. Growth & Differ. 1991; 33: 587-598