1. Trang chủ
  2. » Luận Văn - Báo Cáo

Báo cáo khoa học: Assembly of nuclear DNA-encoded subunits into mitochondrial complex IV, and their preferential integration into supercomplex forms in patient mitochondria doc

13 315 0
Tài liệu đã được kiểm tra trùng lặp

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 13
Dung lượng 553,96 KB

Các công cụ chuyển đổi và chỉnh sửa cho tài liệu này

Nội dung

We also found that Cox6a and Cox7a are incorporated into a novel intermediate complex of approximately 250 kDa, and that transition of subunits from this complex to the mature holoenzyme

Trang 1

mitochondrial complex IV, and their preferential

integration into supercomplex forms in patient

mitochondria

Michael Lazarou1, Stacey M Smith1,2, David R Thorburn2, Michael T Ryan1and

Matthew McKenzie1

1 Department of Biochemistry, La Trobe University, Melbourne, Australia

2 Murdoch Children’s Research Institute and Genetic Health Services Victoria, Royal Children’s Hospital, and Department of Pediatrics, University of Melbourne, Australia

Introduction

The mitochondrial respiratory chain consists of four

multi-subunit complexes (I–IV) that, through

electron-transfer reactions, generate a proton gradient for the

F1Fo-ATPase (complex V) to synthesize ATP

Cyto-chrome c oxidase (complex IV) catalyzes the final step

of the electron transfer chain, in which electrons are

transferred from reduced cytochrome c to molecular oxygen Mammalian complex IV comprises 13 differ-ent subunits, and the crystal structure of the bovine complex was solved over 10 years ago [1] The three largest and most hydrophobic subunits (CO1–3) form the catalytic core, and all are encoded by

mitochon-Keywords

complex IV; cytochrome c oxidase;

membrane proteins; mitochondria; oxidative

phosphorylation

Correspondence

M T Ryan, Department of Biochemistry,

La Trobe University, 3086 Melbourne,

Australia

Fax: +61 3 94792467

Tel: +61 3 94792156

E-mail: m.ryan@latrobe.edu.au

(Received 4 June 2009, revised 18 August

2009, accepted 16 September 2009)

doi:10.1111/j.1742-4658.2009.07384.x

Complex IV is the terminal enzyme of the mitochondrial respiratory chain

In humans, biogenesis of complex IV involves the coordinated assembly of

13 subunits encoded by both mitochondrial and nuclear genomes The early stages of complex IV assembly involving mitochondrial DNA-encoded subunits CO1 and CO2 have been well studied However, the latter stages, during which many of the nuclear DNA-encoded subunits are incorporated, are less well understood Using in vitro import and assembly assays, we found that subunits Cox6a, Cox6b and Cox7a assembled into pre-existing complex IV, while Cox4-1 and Cox6c subunits assembled into subcomplexes that may represent rate-limiting intermediates We also found that Cox6a and Cox7a are incorporated into a novel intermediate complex of approximately 250 kDa, and that transition of subunits from this complex to the mature holoenzyme had stalled in the mitochondria of patients with isolated complex IV deficiency A number of complex IV subunits were also found to integrate into supercomplexes containing combinations of complex I, dimeric complex III and complex IV Subunit assembly into these supercomplexes was also observed in mitochondria of patients in whom monomeric complex IV was selectively reduced We con-clude that newly imported nuclear DNA-encoded subunits can integrate into the complex IV holoenzyme and supercomplex forms by associating with pre-existing subunits and intermediate assembly complexes

Abbreviations

BN-PAGE, blue-native polyacrylamide gel electrophoresis; CAP, chloramphenicol; DDM, n-dodecyl-b- D -maltoside; Dw m, membrane potential, LSI, late-stage intermediate.

Trang 2

drial DNA (mtDNA) The remaining 10 subunits are

encoded by the nuclear genome, and, like other

pro-teins, must be synthesized on cytosolic ribosomes

before being imported and subsequently assembled at

the mitochondrial inner membrane [2] The subunits of

complex IV encoded by nuclear DNA (nDNA) do not

harbor enzymatic activity but function in the structural

integrity, regulation and dimerization of the enzyme

For example, subunits Cox6a and Cox5 play roles in

the regulation of enzymatic activity [3,4], while Cox6b,

along with Cox6a, provides contacts sites for

dimeriza-tion [1,5]

Mammalian complex IV forms a complex of

approx-imately 200 kDa on blue native (BN) PAGE, but is

also found in supercomplexes together with complex I

and dimeric complex III [6] More recently, it has been

suggested that this supercomplex can associate with

complex II, cytochrome c and complex V in a

func-tional respirasome [7] It has been suggested that

supercomplexes enhance respiration due to coordinated

channeling of the electron carriers ubiquinol and

cyto-chrome c [6,8,9]

Isolated complex IV deficiency is one of the most

common respiratory chain defects in humans, and is

associated with various clinical phenotypes such as

mitochondrial encephalomyopathy, lactic acidosis and

stroke-like episodes (MELAS), Leigh disease and lactic

acidosis [10,11] Enzymatic deficiencies of complex IV

that cause disease are often due to defects in the

assembly and⁄ or stability of the enzyme To assemble

complex IV, subunits translated from both genomes

must come together in a coordinated and regulated

manner Studies carried out in the model organism

Saccharomyces cerevisiae have provided insights into

the biogenesis of complex IV, and led to the

identifica-tion of over 20 assembly factors [12] These

nDNA-encoded assembly factors are not associated with

assembled complex IV but instead act at various

func-tional levels, ranging from subunit insertion and

co-factor attachment to regulation of transcription⁄

translation Although yeast COX is a close model of

its human counterpart, significant differences exist For

example, almost half of the yeast assembly factors do

not appear to have human orthologs, and two subunits

(Cox7b and Cox8) are found in the human assembly

but not in yeast In addition, as yeast lacks complex I,

the supercomplex assemblies differ substantially from

those seen in mammalian mitochondria By tracking

the subunit composition of subcomplexes using

trans-lational inhibitors and metabolic labeling of

mtDNA-encoded proteins, a model for the assembly of human

complex IV was proposed [13] This model was later

refined through studies analyzing the composition of

sub-assemblies in the mitochondria of patients deficient

in complex IV [14,15] The model proposes that com-plex IV is assembled via pre-formed intermediates in a stepwise process Of the 10 nDNA-encoded subunits, only Cox4 and Cox5a are integrated at the early stages

of assembly that involve formation of the catalytic core The remaining nDNA-encoded subunits are thought to be added during the last steps of assembly; however, this part of the assembly pathway is the least studied and requires clarification Furthermore, with the model focusing on the de novo synthesis of com-plex IV, it is unclear how newly imported nDNA-encoded subunits are incorporated in the presence of pre-existing holo-complex IV

In this study, we address the assembly of newly imported complex IV subunits in mitochondria from control cells and from cells of patients with defects in complex IV biogenesis We found that, in the presence

of pre-existing complex IV, newly imported nDNA-encoded subunits can integrate into the holoenzyme as well as into its supercomplex forms Furthermore, sub-units Cox6a and Cox7a integrate into a novel late-stage intermediate complex of approximately 250 kDa Assembly into, and progression from, this intermediate was defective in the mitochondria of patients deficient

in complex IV, suggesting that it represents an impor-tant step in assembly of the holoenzyme

Results

In vitro import and assembly of nDNA-encoded complex IV subunits

The assembly of a number of nDNA-encoded com-plex IV subunits was investigated by importing them into isolated mitochondria and monitoring their assem-bly using BN-PAGE As isolated mitochondria are used and protein synthesis does not take place under the conditions used (data not shown), the integration

of a select newly imported subunit into a complex occurs through its association with pre-existing units within the organelle [16,17] Representative sub-units with cleavable (Cox4-1, Cox6a, Cox7a) and non-cleavable (Cox6b and Cox6c) presequences were selected for investigation With the exception of Cox4-1, these nDNA-encoded complex IV subunits are postulated to integrate late in the assembly pathway [13,15] Based on the crystal structure of complex IV [1], the selected subunits are positioned peripherally within the complex (Fig 1A), and, apart from Cox6b, all contain a single transmembrane-spanning domain

35S-labeled complex IV subunit precursor proteins were generated in vitro using rabbit reticulocyte lysate,

Trang 3

and incubated with mitochondria isolated from

cul-tured human fibroblasts for 10 or 60 min in the

pres-ence or abspres-ence of a membrane potential (Dwm) After

import, external proteinase K was added to half of

each sample to degrade non-imported protein Samples

were then subjected to SDS–PAGE, and radiolabeled

subunits were detected using phosphorimage analysis

(Fig 1B) 35S-labeled complex IV subunits bound to

mitochondria, with the signal increasing over time

(Fig 1B, lanes 2 and 3) For the

presequence-contain-ing subunits Cox4-1, Cox6a and Cox7a, an additional

faster-migrating species accumulated, representing the

mature form of the protein (Fig 1B, lanes 2 and 3)

An additional band seen after proteinase K treatment

of Cox7a samples most likely represents a

protease-resistant domain of the precursor, as it is also present

in the absence of import (lane 7) Successful import of

all subunits was determined by their protection from

externally added proteinase K (Fig 1B, lanes 5 and 6),

their dependence on the membrane potential (Dwm) for

this protection (Fig 1B, lanes 4 and 7), and, in the

case of Cox4-1, Cox6a and Cox7a, processing of their

presequences

We next tested the assembly of newly imported

radiolabeled complex IV subunits using BN-PAGE

Radiolabeled subunits were imported into isolated mitochondria and all samples were treated with pro-teinase K before solubilization in n-dodecyl-b-d-malto-side (DDM) and BN-PAGE analysis (Fig 2A) Use of this detergent results in partial dissociation of respira-tory chain supercomplexes, liberating monomeric (holo-) complex IV and a complex III2⁄ complex IV supercomplex [18] The migration of complex IV, dimeric complex III, complex I and their super-complexes (CIII2⁄ CIV and CI ⁄ CIII2) was determined

by western blot analysis (Fig 2A, right panels) Imported Cox6a and Cox7a were incorporated into both monomeric and supercomplex (CIII2⁄ CIV) forms

of complex IV, with the additional presence of an approximately 250 kDa complex (marked with an asterisk) that resolved after 10 min of import (Fig 2A, lanes 3, 4, 9 and 10) Radiolabeled Cox6b also appeared to be incorporated into the holoenzyme, albeit weakly (Fig 2A, lanes 5 and 6), while

35S-labeled Cox4-1 was not incorporated into any distinct complexes, although some high-molecular-weight smearing was evident (lanes 1 and 2) Two dis-tinct complexes ranging between approximately 100 and 150 kDa were seen with newly imported Cox6c (Fig 2A, lanes 7 and 8), although neither of these

A

B

Fig 1 Import of nDNA-encoded complex IV

subunits (A) Structural position of subunits

Cox4 (magenta), Cox6b (red), Cox6c (green),

Cox7a (blue) and Cox6a (orange) within the

crystal structure of bovine complex IV [1].

(B) SDS–PAGE analysis of imported

radiola-beled complex IV subunits Precursor (p)

and mature (m) forms of the subunits are

identified Samples were imported into

mito-chondria in the presence or absence of a

membrane potential (Dwm), and treated with

or without externally added proteinase K

(Prot K) A sample of lysate (representing

20% of added protein ⁄ import) is also shown

(lane 1) Radiolabeled proteins were

detected by phosphorimage analysis.

Trang 4

complexes co-migrated with holo-complex IV Based

on the BN-PAGE analysis of nDNA-encoded

com-plex IV subunits, it can be concluded that some newly

imported subunits can assemble with pre-existing

subunits into holo-complex IV Other subunits do not

integrate into complex IV, perhaps because there is

impaired progression of intermediates along the

assem-bly pathway due to the use of isolated mitochondria

Our results are consistent with those of a previous

study that analyzed the assembly of yeast complex IV

subunits Cox4p, Cox5ap and Cox10p (equivalent to

human Cox5b, Cox4-1 and Cox6a, respectively) [19]

In order to characterize the assembly of

nDNA-encoded complex IV subunits into the various

super-complex forms of super-complex IV, mitochondria were sol-ubilized in digitonin after subunit import In this case, the respiratory chain components are also found in their supercomplex forms [18] The relative positions

of supercomplexes comprising complex I, complex III and complex IV (CI⁄ CIII2⁄ CIV), complex III and complex IV (CIII2⁄ CIV), as well as dimeric com-plex III (CIII2) and monomeric complex IV (CIV), were shown by western blot analysis (Fig 2B, right panels) Radiolabeled Cox6c was incorporated into a complex of approximately 200 kDa that resolved at a slightly lower position than the holo-complex IV, and also into a larger complex at approximately 1 MDa Likewise, 35S-labeled Cox4-1 was found in large

A

CIV

*

CI 669

440

kDa

Cox4-1Cox6aCox6b Cox6c Cox7a

BN-PAGE 0.65% DDM

Time (min)10 60 10 60 10 60 10 60 10 60

α-comple

x I

α-comple

x III

α-comple

x IV

134

67

B

669

440

kDa

Cox4-1Cox6aCox6b Cox6c Cox7a

CIV

BN-PAGE 1% Digitonin

Time (min)10 60 10 60 10 60 10 60 10 60

CIII

α-comple

x I

α-comple

x III

α-comple

x IV

134

67

*

Fig 2 BN-PAGE analysis of imported radio-labeled complex IV subunits 35 S-labeled complex IV subunits were individually incu-bated with isolated fibroblast mitochondria for increasing times as indicated Samples were treated with proteinase K, and solubi-lized in either (A) DDM-containing buffer or (B) digitonin-containing buffer Radiolabeled proteins were detected by phosphorimage analysis Right panels: complex IV (CIV), complex I (CI), complex III (CIII 2 ), and their supercomplex forms (CI ⁄ CIII 2 ), (CI ⁄ CIII 2 ⁄ CIV) and (CIII 2 CIV) were identified by western blot analysis using antibodies to the complex I subunit NDUFA9 (a-com-plex I), the core I subunit of com(a-com-plex III (a-complex III) and the COI subunit of complex IV (a-complex IV) The asterisk indicates a complex of approximately

250 kDa.

Trang 5

complexes in the range of approximately 700–1000

kDa (lanes 1 and 2), but these did not co-migrate with

any of the complex IV-containing supercomplexes The

identity of these complexes is unknown The complex

of approximately 250 kDa observed for newly

imported subunits Cox6a and Cox7a after DDM

solu-bilization was poorly resolved using digitonin (Fig 2B,

marked with an asterisk) However, Cox6a and Cox7a

as well as Cox6b assembled into holo-complex IV and

into complexes co-migrating with the supercomplex

forms CIII2⁄ CIV and CI ⁄ CIII2⁄ CIV Of note, the

intensity of assembled35S-labeled Cox6b was stronger

in mitochondria solubilized in digitonin than those

sol-ubilized in DDM (compare lane 6 in Fig 2A and 2B)

This is consistent with a previous report indicating that

Cox6b can be preferentially stripped from complex IV

in the presence of DDM [20] due to the peripheral

nat-ure of this subunit From these results, we conclude

that, in isolated mitochondria, some newly imported

subunits have the capacity to integrate into the

holoen-zyme as well as into supercomplexes containing

com-plex IV Furthermore, given that yeast mitochondria

lack complex I, this analysis also characterized

nDNA-encoded subunit integration into more complicated

supercomplexes that additionally contain pre-existing

complex I

Assembly profile of imported Cox6a in human

and yeast mitochondria

At early stages of import, the nDNA-encoded

com-plex IV subunits Cox6a and Cox7a were found also to

be incorporated into a complex of approximately

250 kDa, slightly higher than monomeric complex IV

Both subunits are thought to integrate into the

assem-bly pathway at a late stage [13,21] By studying the

import and assembly of Cox6a (see below), we have

established that the complex of approximately

250 kDa represents a novel intermediate, and have

termed it the late-stage intermediate (LSI) complex

(Fig 3) 35S-labeled Cox6a was imported into isolated

mitochondria for various times before solubilization in

DDM and BN-PAGE analysis (Fig 3A) At early time

points, Cox6a was predominantly incorporated into

the LSI complex, while a minor amount accumulated

into holo-complex IV and its supercomplex, CIII2⁄

CIV Over the time course of the experiment, Cox6a

accumulated into holo-complex IV, while the signal for

the LSI complex remained relatively constant Western

blot analysis confirmed the relative positions of

com-plex IV and CIII2⁄ CIV (Fig 3A, right-hand panel);

however, the LSI complex was not seen Assembly

intermediates generally cannot be resolved with

anti-bodies due to their low steady-state levels, but import using small amounts of radiolabeled protein can detect their presence [16,22–24] Next, an in vitro import and chase experiment was performed 35S-labeled Cox6a was imported for 5 min to accumulate the subunit at the LSI complex before mitochondria were re-isolated and then further incubated in the absence of additional radiolabeled precursor (Fig 3B, lanes 1–4) Over the chase period, the intensity of 35S-labeled Cox6a in the LSI complex decreased, with a concomitant increase in the intensity of labeling of holo-complex IV

35S-labeled Cox6a was also imported into mitochon-dria isolated from cells that had been pre-incubated with chloramphenicol (CAP) Under these conditions, complex IV assembly intermediates containing mtDNA-encoded subunits are unlikely to be present, and a pool of unassembled nDNA-encoded subunits may accumulate Any 35S-labeled Cox6a assembly into holo-complex IV is therefore likely to occur via inte-gration into the pre-existing complex as opposed to new assemblies As can be seen in Fig 3C, 35S-labeled Cox6a was incorporated into the LSI complex and holo-complex IV in mitochondria from both control (lanes 1–3) and CAP-treated cells (lanes 4–6) Given that the import of 35S-labeled Cox6a was unaffected (Fig 3D), the decreased signal of assembled Cox6a observed in the CAP-treated samples is probably a result of decreased levels of fully assembled com-plex IV as shown by western blot analysis (bottom panels in Fig 3C) As in organello labeling is inefficient under the conditions used (data not shown) and Cox6a assembly occurs even in mitochondria isolated from CAP-pretreated cells, these results support the possibil-ity that late-assembling subunits such as Cox6a have the capacity to assemble into complex IV by cycling with pre-existing subunits

In order to eliminate the possibility that the precur-sor form is found in the LSI complex, 35S-labeled Cox6a was imported into mitochondria for various times (without proteinase K treatment) and subjected

to BN-PAGE followed by SDS–PAGE in the second dimension (Fig 3E) At early time points, the precur-sor form of 35S-labeled Cox6a (as judged by its co-migration with the lysate control sample) was found

in a high-molecular-weight smear, presumably bound

to molecular chaperones and⁄ or the translocase of the outer mitochondrial membrane (TOM) machinery [25] The mature form of 35S-labeled Cox6a was initially found in the more slowly migrating LSI complex, and over time assembled into mature complex IV The position of mature complex IV was confirmed by immunostaining (bottom panel) Based on these results, we conclude that the LSI complex represents a

Trang 6

novel intermediate for the biogenesis of Cox6a, and

hence Cox7a, into human complex IV Coomassie

staining of mitochondria did not reveal the presence of

the LSI complex even though complex IV could be

detected (data not shown), consistent with the LSI

complex being a short-lived intermediate

As an ortholog of Cox6a is found in yeast [termed

Cox10p (COX13)], we determined whether the LSI

complex is evolutionarily conserved In vitro import of

Cox10p has previously been used to analyze

com-plex IV assembly in yeast, but digitonin-solubilized

mitochondria were employed and hence only

super-complexes were observed [19] Furthermore, a time

course of the import was not performed Radiolabeled Cox10p was imported into isolated yeast mitochondria, and all samples were treated with proteinase K before solubilization in DDM and BN-PAGE analysis (Fig 4A) The time-course analysis revealed that the assembly pathway of Cox10p resembles that of its human counterpart The intensity of the LSI complex remained consistent over time, and the holo-com-plex IV of approximately 200 kDa accumulated Import and chase analysis of 35S-labeled Cox10p (Fig 4B, lanes 1–4), revealed that, like its human Cox6a counterpart, Cox10p initially accumulated in the LSI complex and was chased over time into holo-complex IV Evolutionary conservation of the LSI complex containing Cox6a⁄ Cox10p suggests that it is

an important step in the biogenesis of these subunits and subsequently the holoenzyme

Assembly analysis of35S-labeled Cox6a in patient mitochondria

The assembly of subunit Cox6a was investigated fur-ther using fibroblast mitochondria from two patients

LSIC CIV

669 440 kDa

α-complex IV

669

440

kDa

LSIC CIII2/CIV

CIII2/CIV

CIII2/CIV

CIV

α-complex I V

35 S-Cox6a (Human)

Time (min)10 30 601 2 3 Chase (min) 0 15 30 601 2 3 4

134

67

134 67

35 S-Cox6a (Human)

669

440

kDa

134

67

LSIC CIV Control + CAP

1 2 3 4 5 6

Time (min) 10 30 60 10 30 60

1 2 3 4 5 6

α -complex IV

α -complex II CIV

CII

Control

+ CAP

1 2 3 4 5 6 Time (min) 10 30 60 10 30 60

p m

p m

- Prot K + Prot K

CIV

BN-PAGE

Lysate

control

5 min

10 min

60 min

30 min

α -complex IV

E

Fig 3 Import and assembly of Cox6a into pre-existing complex IV (A) 35S-labeled Cox6a was incubated for various times with mito-chondria isolated from human fibroblasts Samples were treated with proteinase K and subjected to DDM solubilization, BN-PAGE and phosphorimaging Right lane: the migration of CIV and CIII 2 ⁄ CIV supercomplexes was identified by western blot analysis using antibodies to the complex IV subunit COI (a-complex IV) (B)

35

S-labeled Cox6a was incubated for 5 min with mitochondria after removal of free 35 S-labeled Cox6a and chase of assembly for vari-ous times as indicated Mitochondria were treated as in (A) (C)

35 S-labeled Cox6a was incubated for 10–60 min with mitochondria isolated from control fibroblasts that had been pre-treated with or without chloramphenicol (CAP) for 12 h Samples were treated with proteinase K before being solubilized in DDM-containing buffer and subjected to BN-PAGE, western transfer and phosphorimage analy-sis (top panel), followed by immunodecoration using antibodies to COI (a-complex IV) and 70 kDa subunit (a-complex II) (bottom panel) (D) 35S-labeled Cox6a was imported into mitochondria as described in (C), with and without proteinase K treatment, before SDS–PAGE and phosphorimage analysis (E) 35 S-labeled Cox6a was imported into control mitochondria for increasing times as indi-cated Samples were solubilized in DDM-containing buffer, and sub-jected to BN-PAGE in the first dimension followed by SDS–PAGE

in the second dimension Gels were subjected to phosphorimaging The position of complex IV was confirmed based on immunoblot analysis using antibodies against COI (bottom panel) The lysate control shows the position of the 35 S-labeled Cox6a precursor spe-cies after one-dimensional SDS–PAGE p,35S-labeled Cox6a precur-sor form; m, 35 S-labeled Cox6a mature form; CIV, complex IV; LSIC, late-stage intermediate complex; CIII 2 ⁄ CIV, complex III 2 ⁄ com-plex IV supercomcom-plex.

Trang 7

with Leigh syndrome involving isolated complex IV

deficiency [18] Patient 1 was homozygous for a

patho-genic frameshift mutation in the gene encoding the

complex IV assembly factor Surf1 Patient 2 had

iso-lated complex IV deficiency and an as yet unknown

nuclear gene mutation, although mutations in

com-plex IV subunits and known assembly factors have

been excluded (data not shown) This analysis served

to test whether defects in formation of the LSI

com-plex may be involved in altered comcom-plex IV assembly

and human disease, and also to assess the possible involvement of Surf1 in the formation and assembly of the LSI complex BN-PAGE and western blot analysis

of mitochondria solubilized in DDM indicated that the fibroblasts of patients 1 and 2 had very low levels of mature complex IV (Fig 5A, right panels) In addi-tion, the CIII2⁄ CIV supercomplex was not detected in patient mitochondria, although the levels of complex I, complex III and the CI⁄ CIII2supercomplex were simi-lar to control The amount of assembled 35S-labeled Cox6a in the mitochondria of patient 1 (Fig 5A, lanes 5–7) was lower relative to control (lanes 1–3), with reduced levels of both the LSI complex and holo-com-plex IV This is most likely due to the substantially decreased levels of complex IV in these mitochondria

In the mitochondria of patient 2, 35S-labeled Cox6a was initially incorporated into the LSI complex (Fig 5A, lane 13) where it accumulated over time (lanes 14 and 15) The radioactivity below the LSI complex and at the position of the CIII2⁄ CIV complex may represent low levels of assembled complex IV (lane 15) Thus, it appears that progression from the LSI complex to holo-complex IV is defective in the mitochondria of patient 2 compared to the assembly profile for the control (compare lanes 9–11 and 13–15)

As expected, disruption of the membrane potential (Dwm) abolished assembly in all samples due to inhibi-tion of import SDS–PAGE was also performed in order to eliminate the possibility that the assembly defects for Cox6a in patient mitochondria were a result

of impaired import As shown in Fig 5B, 35S-labeled Cox6a was efficiently imported into both patient and control mitochondria in a Dwm-dependent manner

Time (min)10 30 60 10 30 60

Patient 1 Control

60 60

CIII

2 /CIV CIV LSIC

669

440

kDa

10 30 60 10 30 60

Patient 2 Control

60 60

+ + + + + + –

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

α -complex IV α -complex III α -complex I

CIII2 CI/CIII2

Time (min)10 30 60 60 10 30 60 60

SDS-PAGE

Control

Patient 1

p m

p m –Prot K +Prot K

Lysate

Patient 2 p

m

1 2 3 4 5 6 7 8 9

BN-PAGE 0.65% DDM

Control Patient 1 Patient 2 Contro

l

Patient 1 Pati ent 2 Contro l Patient 1 Patient 2

134

67

Fig 5 Import and assembly of Cox6a in control and patient mitochondria Mitochondria from control or patient cells were incubated with

35

S-labeled Cox6a for increasing times in the presence or absence of a membrane potential (Dw m ) Half of each sample was treated with proteinase K before being split in two and (A) solubilized in DDM-containing buffer and subjected to BN-PAGE (protease-treated samples only), or (B) SDS–PAGE and phosphorimaging The right panels in (A) show western blot analysis of complex IV, complex III and complex I

in mitochondrial preparations CI, complex I; CIII 2 , complex III dimer; CIV, complex IV; LSIC, late-stage intermediate complex; CIII 2 ⁄ CIV, complex III 2 ⁄ complex IV supercomplex; CI ⁄ CIII 2 , complex I ⁄ complex III supercomplex.

669

440

kDa

Time (min) 10 30 60

LSIC CIV

134

67

35S-Cox10p (Yeast)

669

440 kDa

Chase (min) 0 15 30 60

35S-Cox10p (Yeast)

LSIC CIV

134 67

Fig 4 Import and assembly of the yeast Cox6a ortholog Cox10p

in yeast mitochondria (A) 35 S-labeled Cox6a was incubated for

vari-ous times with mitochondria isolated from yeast cells Samples

were treated with proteinase K and subjected to DDM

solubiliza-tion, BN-PAGE and phosphorimaging (B) 35 S-labeled Cox6a was

incubated for 5 min with mitochondria after removal of free

35 S-labeled Cox6a and chase of assembly for various times

Mito-chondria were treated as in (A) CIV, complex IV; LSIC, late-stage

intermediate complex.

Trang 8

These results suggest that incorporation of Cox6a into

the LSI complex and its progression to complex IV is

not dependent on Surf1, and the slowed rate of

assem-bly into complex IV is most likely a result of reduced

levels of holo-complex IV In the mitochondria of

patient 2, the progression of Cox6a into complex IV is

effectively stalled, suggesting that the underlying defect

may be related to a late stage in complex IV biogenesis

Assembly of nDNA-encoded complex IV subunits

into supercomplexes in control and patient

mitochondria

When we analyzed respiratory complexes in patient

mitochondria after digitonin solubilization and

BN-PAGE, we found that the residual complex IV was predominantly found in supercomplexes Using anti-bodies against the subunits of complexes I, III and IV (Fig 6A, right panels), it appeared that only the super-complex form (CI⁄ CIII2⁄ CIV) was present in patient mitochondria An additional faster-migrating species was also detected in patient mitochondria, and is likely

to represent the CI⁄ CIII2 supercomplex In control mitochondria, complex IV was also found in its super-complex forms, but its most predominant form was as

a monomer As complex IV was only detected in the

CI⁄ CIII2⁄ CIV supercomplex in patient mitochondria, this suggests that this form is particularly stable and⁄ or crucial for respiratory function However, we cannot exclude the possibility that complex IV found

Cox4-1 Cox6b Cox6c Cox7a Cox6a

Time (min) 10 60 10 60 10 60 10 60 10 60

669 440

kDa

CI/CIII2/CIV

CIV CIII2

BN-PAGE 1% Digitonin

α-complex I

α-complex III

α-complex IV

CI/CIII2

669

440

kDa

A

B

BN-PAGE 1% Digitonin α -complex IV α -complex III α -complex I

Control Patient 1 Patient 2

CIII2/CIV

CIV CIII2

CI/CIII2/CIV

CI/CIII2 CI/CIII2/CIV

Control Patient 1 Patient 2 Control Patient 1 Patient 2 Time (min)10 30 60 10 30 60

Patient 1 Control

60 60

+

Patient 2 Control

60 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

134

67

134

67

*

LSIC

Fig 6 nDNA-encoded subunits assemble into supercomplexes of control and patient mitochondria (A) Mitochondria from control or patient cells were incubated with 35 S-labeled Cox6a for increasing times in the presence or absence of a membrane potential (Dwm) Each sample was treated with proteinase K before being solubilized in digitonin-containing buffer and subjected to BN-PAGE and phosphorimaging (B)

35 S-labeled complex IV subunits were incubated with mitochondria from patient 1 for 10 and 60 min and treated as described in (A) The right panels in (A) and (B) show the migration of complex IV (CIV), dimeric complex III (CIII2), CIII2⁄ CIV supercomplex, complex I (CI) ⁄ CIII 2 ⁄ CIV supercomplex and CI ⁄ CIII 2 supercomplex by western blot analysis The asterisk indicates the complex of approximately 100 kDa.

Trang 9

in supercomplexes of patient mitochondria is not all

fully assembled

We determined whether the integration of newly

imported Cox6a into supercomplexes was defective in

patient mitochondria Radiolabeled Cox6a was

imported into mitochondria isolated from patient 1,

patient 2 and control fibroblasts, solubilized in

digito-nin and analyzed using BN-PAGE (Fig 6A) As

observed above (Fig 5), the assembly of newly

imported Cox6a into monomeric complex IV was

reduced in mitochondria from cells of both patient 1

(Fig 6A, lanes 5–7) and patient 2 (lanes 13–15)

Although the LSI complex was not clearly resolved

after solubilization using digitonin, mitochondria from

cells of patient 2 showed smearing above complex IV

(lanes 13–15), and this may represent the impaired

pro-gression of35S-labeled Cox6a from the LSI complex to

holo-complex IV Incorporation of 35S-labeled Cox6a

into the CIII2⁄ CIV supercomplex was reduced in

mito-chondria from both patient 1 and patient 2; however,

its incorporation into the CI⁄ CIII2⁄ CIV supercomplex

was not impaired, with an overall signal comparable to

that in controls (Fig 6A)

The relatively efficient assembly of newly imported

Cox6a into the CI⁄ CIII2⁄ CIV supercomplex in patient

mitochondria led us to investigate the assembly of

additional nDNA-encoded subunits Mitochondria

iso-lated from fibroblasts from patient 1 were chosen for

this study because of the better growth rate of these

cells in culture As shown in Fig 6B, newly imported

Cox4-1, Cox6b, Cox6c, Cox7a and Cox6a all

effi-ciently assembled into the CI⁄ CIII2⁄ CIV supercomplex

in the mitochondria of patient 1 The subunit assembly

profile also differed to that observed in control

mito-chondria, in that Cox4-1 and Cox6c did not assemble

into any complex IV-containing supercomplexes (see

Fig 2B, lanes 1–2 and lanes 5–6) However, in

addi-tion to its assembly into supercomplexes in the

mito-chondria of patient 1, Cox4-1 also assembled into a

complex of approximately 100 kDa (Fig 6B, lanes 1

and 2) that was not observed in control mitochondria

Cox4-1 is believed to form an early assembly

interme-diate with CO1 [15], and the species of approximately

100 kDa may represent such an intermediate; however,

further characterization is required Additional

low-molecular-weight complexes were also seen for Cox6c

import in the mitochondria of patient 1 that were not

seen in control mitochondria (see Fig 2B, lanes 5 and

6) These smaller complexes may represent

rate-limit-ing intermediates due to assembly defects in the

mito-chondria of patient 1 It is possible that, in the absence

of monomeric complex IV, a portion of these

inter-mediates can combine with complex I and complex IV

to form the supercomplexes that are seen under these conditions

Discussion

The current model of complex IV assembly follows a sequential pathway that begins with the mitochondrial translation of CO1 followed by integration of addi-tional subunits and co-factors through a set of defined intermediates (for reviews on complex IV assembly, see [12,21,26,27]) A number of assembly factors have been identified that assist in the process and act at the levels

of regulation [28], co-factor biosynthesis and insertion [29,30] and chaperoning of assembly intermediates [31] Much of our current knowledge regarding complex IV biogenesis has been provided by studies using the model organism S cerevisiae Most studies have focused on the early stages of assembly that involve formation of the catalytic core consisting of the mtDNA-encoded subunits CO1, CO2 and CO3 Thus details of the latter stages in which the majority of nDNA-encoded subunits that surround the core are assembled remain largely unknown In particular, it is not clear how a newly imported nDNA-encoded subunit assembles in the presence of pre-existing complex IV Furthermore, there are a number of dif-ferences between mammalian and yeast mitochondria that affect complex IV biogenesis, and these may be relevant to disease These include the presence of struc-tural subunits and assembly factors in yeast that do not appear to have homologs in mammals, and differences in supercomplex forms The consequence of these differences is that further analysis of the biogene-sis of the mammalian enzyme is required

Assembly of nDNA-encoded complex IV subunits

Of the five nDNA-encoded subunits investigated in this study, Cox6a, Cox7a and Cox6b were found to assemble into both monomeric and supercomplex forms of complex IV (Fig 1) According to the current model of complex IV assembly, subunit Cox6b assem-bles into the S3 subcomplex together with a host of other nDNA-encoded subunits as well as CO3 [13,15] However, it has been suggested more recently that Cox6b is incorporated at the very last step of com-plex IV assembly [32], possibly after addition of the late-assembling subunits Cox6a and Cox7a Based on the results presented here, it appears that subunits that are incorporated late in the assembly pathway have a greater propensity to assemble into pre-existing com-plex IV in isolated mitochondria Previously, we have shown that individual, newly imported complex I

Trang 10

subunits can dynamically exchange with their

pre-exist-ing counterparts to assemble into the complex [16]

Similar findings have also been reported for other

complexes [24,33–36] It is therefore possible that these

newly imported nDNA-encoded complex IV subunits

are assembling into the pre-existing complex via a

similar mechanism

The remaining two subunits, Cox4-1 and Cox6c,

were not found to assemble into holo-complex IV;

however, they were found to integrate into additional

complexes Subunit Cox6c resolved into two distinct

complexes in the range of approximately 100–150 kDa

when the detergent DDM was used The assembly

pro-file of Cox6c differed in digitonin-solubilized samples,

where it was found to assemble into a complex of

approximately 160 kDa and also a large species of

approximately 1 MDa Import of subunit Cox4-1

revealed that it also assembled into large complexes

ranging from approximately 700–1000 kDa, although

it was not found in any distinct complexes when DDM

was used This subunit is considered to be one of the

first nDNA-encoded subunits to integrate into

sub-assemblies of complex IV [13,14] In a previous study,

import analysis of the yeast ortholog of human Cox4-1

revealed that it assembles into a number of complexes

ranging from 250 to 450 kDa in size [19] These

com-plexes were found to contain assembly factors such as

Cox14p, Coa1p and Shy1p as well as complex IV

su-bunits Of these assembly factors, only Shy1p has a

human homolog, termed Surf1 [37,38], and this could

account for the different complexes observed in yeast

and mammalian mitochondria Nevertheless, the

com-plexes identified here may represent rate-limiting

inter-mediates that require additional factors not found in

yeast mitochondria for Cox4-1 and Cox6c subunit

maturation and⁄ or assembly

A novel late-stage intermediate complex for

Cox6a and Cox7a assembly into complex IV

In mitochondria solubilized with DDM, newly

imported subunits Cox7a and Cox6a were found to

integrate into the LSI complex of approximately

250 kDa prior to their incorporation into

holo-com-plex IV and its supercomholo-com-plex forms Similar results

were observed in yeast mitochondria when the

assem-bly of the Cox6a ortholog, Cox10p was analyzed, thus

indicating that the assembly profile for this subunit is

evolutionarily conserved In comparison to

com-plex IV, the LSI comcom-plex is not visible on

Coomassie-stained 2D gels (data not shown), supporting the

conclusion that it is a low-level intermediate complex

that is only detected with radiolabeled proteins

Previous import and assembly analysis of Cox10p using digitonin solubilization did not reveal the pres-ence of a complex of approximately 250 kDa [19], con-sistent with our findings that this complex is not well resolved in digitonin Analysis of mitochondria from a patient harboring an isolated complex IV deficiency of unknown etiology (patient 2) revealed that Cox6a assembly into the LSI complex had stalled Although a complex of similar size has previously been described

in yeast mitochondria [19,39], this complex is likely to differ as yeast contains assembly factors that are not present in humans Furthermore, the studies in yeast used digitonin for solubilization, and the LSI complex

is not clearly resolved in this detergent Therefore, we conclude that the LSI complex represents a novel com-plex for maturation of at least Cox6a and Cox7a into complex IV, and that this assembly process is per-turbed in human disease A possible explanation for the larger size of the LSI complex relative to the holo-enzyme is that a specific accessory factor is associated with this late-stage intermediate that is displaced after integration of the subunits into the final complex Alternatively, Cox6a and Cox7a may integrate into the LSI complex and then be transferred into the maturing complex IV

Integration of nDNA-encoded subunits into supercomplexes in patient mitochondria Using digitonin solubilization of mitochondria to visu-alize supercomplexes, it was found that subunit Cox6a assembled into the CI⁄ CIII2⁄ CIV supercomplex in the mitochondria of patients 1 and 2 Immunoblot analysis

of patient mitochondria revealed that all detectable complex IV was present in a supercomplex with com-plexes I and III This is in contrast to control mito-chondria in which the majority of complex IV is not associated with supercomplexes and instead resolves in its monomeric form [18,40,41] As that complex IV may be important for the assembly⁄ stability of com-plex I [42,43] and functions within respirasomes [7], it follows that limiting amounts of complex IV (in partial

or fully assembled forms) could be sequestered within supercomplexes, as observed in patient mitochondria All subunits tested here (Cox4-1, Cox6a, Cox6b, Cox6c and Cox7a) were found to efficiently assemble into the CI⁄ CIII2⁄ CIV supercomplex Of particular interest were the newly imported subunits Cox4-1 and Cox6c, as they did not assemble into complex IV or its supercomplex forms in control mitochondria, but had the ability to integrate into the CI⁄ CIII2⁄ CIV super-complex in patient mitochondria As the subunits Cox4-1, Cox6b, Cox6c, Cox7a, and Cox6a all

Ngày đăng: 07/03/2014, 00:20

TÀI LIỆU CÙNG NGƯỜI DÙNG

TÀI LIỆU LIÊN QUAN

🧩 Sản phẩm bạn có thể quan tâm