Ubiquitin-like protein 5 positively regulates chaperone gene expression in the mitochondrial unfolded protein response

Abstract

Perturbation of the protein-folding environment in the mitochondrial matrix selectively upregulates the expression of nuclear genes encoding mitochondrial chaperones. To identify components of the signal transduction pathway(s) mediating this mitochondrial unfolded protein response (UPR(mt)), we first isolated a temperature-sensitive mutation (zc32) that conditionally activates the UPR(mt) in C. elegans and subsequently searched for suppressors by systematic inactivation of genes. RNAi of ubl-5, a gene encoding a ubiquitin-like protein, suppresses activation of the UPR(mt) markers hsp-60::gfp and hsp-6::gfp by the zc32 mutation and by other manipulations that promote mitochondrial protein misfolding. ubl-5 (RNAi) inhibits the induction of endogenous mitochondrial chaperone encoding genes hsp-60 and hsp-6 and compromises the ability of animals to cope with mitochondrial stress. Mitochondrial morphology and assembly of multi-subunit mitochondrial complexes of biotinylated proteins are also perturbed in ubl-5(RNAi) worms, indicating that UBL-5 also counteracts physiological levels of mitochondrial stress. Induction of mitochondrial stress promotes accumulation of GFP-tagged UBL-5 in nuclei of transgenic worms, suggesting that UBL-5 effects a nuclear step required for mounting a response to the threat of mitochondrial protein misfolding.

Figure 1.—

(A) Hypothesized components of a mitochondrial UPR. Physiological and developmental cues impose an unfolded protein load on the mitochondria. The resultant physiological stress activates the afferent limb of the UPR^mt, increasing expression of genes encoding mitochondrial chaperones. The latter serve as the pathway's efferent limb that restores homeostasis to the organelle. A mutation such zc32 (see below), which activates the afferent limb by causing more mitochondrial stress, may achieve this by compromising the UPR^mt's efferent limb or by imposing an increased burden of unfolded/misfolded proteins. A gene whose inactivation diminishes the activity of the afferent limb (ubl-5, see below) may function in propagating the stress signal or may suppress the pathway upstream by diminishing the load of unfolded/misfolded proteins that enter the organelle. (B) Activation of the UPR^mt by a temperature-sensitive mutation, zc32. Shown are fluorescent micrographs of animals transgenic for a GFP reporter linked to the UPR^mt-inducible promoters of the mitochondrial chaperone-encoding genes hsp-6 and hsp-60. Developing wild-type (N2) embryos were exposed to spg-7(RNAi) or to ethidium bromide (Et Br), which induce the UPR^mt. Mutant zc32 embryos were allowed to develop to adulthood at the permissive or nonpermissive temperature. (C) Photomicrographs of progeny of wild-type (N2) and zc32 mutant adults that had developed at the permissive (20°) or nonpermissive (25°) temperature. The mean (±SEM) number of progeny per mutant hermaphrodite reaching the L4 stage is indicated below the photomicrograph.

Figure 2.—

The zc32 mutation compromises the ability to cope with further mitochondrial stress. Shown are photomicrographs of plates with wild-type (N2) and zc32 mutant animals that developed at the permissive (20°) and nonpermissive temperature (25°) from embryos exposed to mock RNAi or RNAi of genes that induce further mitochondrial stress (spg-7, phb-2, and hsp-60) or ER stress (ire-1). Photographs were taken 84 hr after seeding the RNAi plates with embryonated eggs. The fraction of animals that reached developmental stage ≥L4 (mean ±SEM) is indicated below the images.

Figure 3.—

Inactivation of ubl-5 compromises signaling in the UPR^mt. (A) Fluorescent photomicrographs of wild-type (N2) and zc32 mutant animals transgenic for the UPR^mt reporters hsp-6∷gfp and hsp-60∷gfp or for the UPR^er reporter hsp-4∷gfp. Animals developed from founders placed on mock RNAi or ubl-5(RNAi) plates at 20° and were shifted, where indicated, to the nonpermissive temperature (25°) for 24 hr before analysis. The animals in the left panels were exposed to genomic-based RNAi whereas those on the right were exposed to cDNA-based RNAi constructs derived from C. elegans mRNA (Ce) or C. briggsae mRNA (Cb). Where indicated, the hsp-4∷gfp reporter was activated by exposure to the ER-stress-inducing drug tunicamycin [note that ubl-5(RNAi) does not affect the induction of hsp-4∷gfp]. (B) Photomicrographs of progeny of wild-type (N2) or zc32 founder animals that developed on mock or ubl-5(RNAi) plates maintained for 16 hr at the permissive temperature of 20° (to allow egg laying by the founder) and shifted to the nonpermissive temperature of 25° for 3 days. The number of F₁ progeny that reached developmental stage ≥L4 (mean ±SEM)/F₀ hermaphrodite is indicated below the images. (C) Immunoblot of GFP reporter in hsp-60∷gfp transgenic wild-type, zc32 mutant, or animals exposed to the indicated mitochondrial stress-inducing RNAi in the presence or absence of ubl-5(RNAi). The anti-HDEL blot, which detects C. elegans BiP, serves as a loading control. (D) Northern blot of endogenous hsp-60 and hsp-6 RNA in animals that developed on mock RNAi, spg-7(RNAi) (to induce mitochondrial stress), and a combination of spg-7(RNAi) and ubl-5(RNAi). Samples were processed 72 and 96 hr after seeding the RNAi plates with eggs. Ribosomal RNAs stained with ethidium bromide serve as a loading control. (E) Immunoblot of GFP, as in C. The GFP protein encoded by the hsp-60∷gfp(zcIs9) transgenic UPR^mt reporter (hsp-60∷GFP) and the UBL-5Cb∷GFP fusion protein encoded by the rescuing transgenic allele of ubl-5^Cb∷gfp(zcIs22) are indicated. The asterisk marks an immunoreactive protein fragment found in ubl-5^Cb∷gfp(zcIs22) animals, which is likely an in vitro proteolytic fragment of the fusion protein. The anti-HDEL blot, which detects C. elegans BiP, serves as a loading control.

Figure 4.—

Inactivation of ubl-5 selectively compromises an animal's ability to cope with mitochondrial stress. (A) High-magnification fluorescent photomicrographs of mitochondria labeled with a GFP with a cleavable mitochondrial import signal (GFP^mt) expressed from a myo-3∷gfp^mt transgene. Animals developed to adulthood from embryos exposed to the indicated RNAi. Note the irregular morphology of the mitochondria in animals exposed to RNAi that induces mitochondrial stress (spg-7, hsp-60, phb-2) or inactivates ubl-5 and the normal morphology of the mock and ire-1(RNAi) animals. (B) Photomicrographs of plates populated by progeny (F₁) of myo-3∷gfp^mt or myo-3∷gfp^cyt animals subjected to the indicated RNAi. The F₀ animals had been removed from the plate. The number of F₁ progeny that reached a developmental stage ≥L4 (mean ±SEM)/F₀ hermaphrodite (n = 4) is indicated below the images. The panels at the far right are fluorescent photomicrographs of transgenic animals. (C) Photomicrographs of progeny produced by the F₁ generation of mock and ubl-5(RNAi) transgenic animals expressing GFP in mitochondria (ges-1∷gfp^mt) or cytoplasm (ges-1∷gfp^cyt) of the intestinal cells. (D) Immunoblot of GFP and the anti-HDEL loading control in test and control pairs of myo-3∷gfp^mt and myo-3∷gfp^cyt or ges-1∷gfp^mt and ges-1∷gfp^cyt transgenic animals.

Figure 5.—

Inactivation of ubl-5 perturbs formation of protein complexes in the mitochondria. (A) Detection of an ∼75-kDa biotinylated protein(s) in fractionated worm extracts by strepavidin–HRP ligand blotting. Total worm extract (T), postmitochondrial supernatant (SN), and mitochondrial pellet (MT) are shown. GFP^cyt expressed from the myo-3 promoter serves as a marker for a cytoplasmic protein. (B) Distribution of an ∼75-kDa biotinylated mitochondrial protein(s) in fractions of glycerol gradient prepared from animals exposed to mock RNAi, hsp-60(RNAi), and ubl-5(RNAi). (Top) Unstressed animals. (Bottom) Mitochondrially stressed myo-3∷gfp^mt animals. The migration of complexes of the indicated size in this gradient is shown. The final fraction (14) also contains the pellet.

Figure 6.—

The stress of protein misfolding in the mitochondrial matrix promotes nuclear localization of UBL-5∷GFP. (A) Fluorescent photomicrographs of ubl-5∷gfp transgenic animals that developed from larvae while exposed to the indicated RNAi. Shown are low magnification (top) and high-magnification (bottom) fluorescent micrographs of the GFP channel and the H33258 nuclear stain. Note that the large polyploid intestinal nuclei stain brightly with H33258 in all samples and with UBL-5∷GFP in the stressed samples. (B) The relative intensity of the cytosolic and nuclear UBL–GFP signal in animals subjected to the indicated genetic manipulations. The mean ±SEM ratio of signal from the nuclear and cytosolic regions of the intestinal cell is plotted (N = 20, * P < 0.05, unpaired two-tailed Student's t-test, compared with mock sample).