| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Departament de Genètica, Universitat de València, Burjassot, Valencia, Spain;
Institute of Zoology, University of Regensburg, Regensburg, Germany; and
Developmental Biology Program, Memorial Sloan Kettering Cancer Institute, New York, New York, USA
2Correspondence: Departament de Genètica, Facultat de Ciències Biològiques, Universitat de València, Carrer Doctor Moliner 50, 46100-Burjassot, Valencia, Spain. E-mail: dmolto{at}uv.es
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: frataxin • aconitase • mitochondrial respiration • hyperoxia • RNAi
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. Extraneural organs are also affected during the course of the disease, as a significant proportion of patients develop cardiomyopathy (1
, 3)
. This disabling condition manifests usually in childhood or adolescence. Patients develop progressive ataxia of all four limbs, dysarthria, sensory loss, and pyramidal signs (1)
. Consequently, they have a diminished life quality, resulting in confinement to a wheel chair and a reduced life expectancy mostly due to the hypertrophic cardiomyopathy.
FA is caused by reduction of the frataxin protein mostly due to an abnormal GAA repeat expansion in the first intron of the human FRDA gene (4)
, which, in turn, inhibits transcription (5)
. Frataxin shows high conservation throughout evolution, with orthologs in essentially all eukaryotes and some prokaryotes (6)
. Its presence in Gram-negative bacteria supports the mitochondrial localization found in eukaryotes (7
, 8)
. Considerable effort has been made to explain the primary dysfunction in the FA pathogenesis. Frataxin has been proposed to play several different roles: regulating efflux of iron from mitochondria (9)
, storing iron in bioavailable and nontoxic form within mitochondria (10
11
12)
, regulating OXPHOS (13
, 14)
, producing iron-sulfur (Fe-S) clusters (15)
; controlling heme group synthesis (16)
, and modulating mitochondrial aconitase activity (17)
.
Oxidative stress has also been suggested to have an important role in the pathophysiology of FA (18)
. High levels of oxidative stress markers in FA patient samples, such as malondialdehyde in plasma (19)
, urinary-excreted oxidized DNA (20)
, and glutathione in blood (21)
, have been found. Studies carried out in different models of FA, including yeast, cell culture, and mouse, have reported an alteration in intracellular oxidative status (22
23
24
25)
. In addition, overexpression of frataxin led to an up-regulation of some antioxidant pathways (26
, 27)
. However, recent studies have questioned the role of oxidative stress in FA. In those reports, the overexpression of antioxidant enzymes as catalase, Cu,Zn-superoxide dismutase, and Mn-superoxide dismutase or MnTBAP treatment (a compound mimicking the Mn-superoxide dismutase action) failed to rescue the loss of frataxin function phenotypes in both mouse and fly (28
, 29)
. Furthermore, Seznec et al. (28)
did not detect increased level of oxidative stress markers, suggesting that oxidative stress may not be a major contributor to the pathology.
To assess both the roles of frataxin and oxidative stress in FA, we induced posttranscriptional silencing of the Drosophila frataxin gene (fh) by transgenic double-stranded RNA interference (RNAi). We have generated a scenario where fh levels are reduced to 30% compared to the controls, thus overcoming the preadult lethality observed in other fly models (29)
. This allowed us to study the effects of Drosophila frataxin (FH) reduction and the contribution of oxidative insult in adult individuals. Furthermore, the effect of general and tissue-specific overexpression of fh was also analyzed.
In this work, we show that FH plays an essential role in Drosophila melanogaster protecting against the deleterious effects of oxidative stress. We have also confirmed the prediction concerning the mitochondrial localization of FH (30)
. Both frataxin-deficient and frataxin-overexpressing adult flies showed reduced life span and climbing ability. Moreover, both the reduction and the increase of FH function seriously compromised aconitase activity in an unbalanced redox environment. FH decrease also impaired mitochondrial respiration in an oxidative atmosphere. This work presents the first evidence regarding the essential function of frataxin in protecting aconitase from oxidative stress-dependent inactivation in a multicellular organism and supports an important role of oxidative stress in the progression of Friedreich ataxia.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Generation of the fh constructs
FH-enhanced GFP construct
The coding sequence of fh was amplified from poly (A)-RNA isolated from Drosophila embryos. The gene-specific primers used were MECAD (5'-AAGTTGCGGCCGCCGCAACTGGGATTTGTA-3') and MEDAR (5'-ACTAATTCTAGAATTAACTACAGTAGGGCA-3'). The resulting polymerase chain reaction (PCR) product was cloned into pCRscript SK(+) to yield the pCR-fh construct. The fh coding region was amplified from this construct using the fhpEGFPN3F (5'-CTCGAGAAATGTTTGCCGGTCGTTTGAT-3') and fhpEGFPN3R (5'-GGATCCACTACAGTAGGGCAGGCGTAGGAAG-3') primers and subcloned in frame with the enhanced GFP (EGFP) protein into pEGFPN3 vector (BD Biosciences Clontech, Mountain View, CA).
UAS-fhIR construct
This construct was generated according to Piccin et al. (31)
. The pCRfh construct was used to obtain two copies of fh in opposite directions. Both copies were separated using a fragment of the green fluorescent protein (GFP) as spacer. This fragment was amplified with the primers GFPRNAiF (5'-CCCAAGCTTCACGAATTCTTCAAGTCCGCC-3') and GFPRNAiR (5'-CCGCTCGAGCTGGATCCGGACTTGTACAGC-3'). Finally, the construct containing the two fh sequences in opposite orientation and separated by the GFP spacer was cloned into the pUAST vector (UAS-fhIR).
UAS-fh construct
The fh coding sequence was subcloned from the pCR-fh construct into the appropriate restriction sites in the polylinker of the pUAST vector (UAS-fh).
Cellular transfections
Transient tranfections were carried out in CHO-K1 mammalian cells. The day before transfections, sterile cover slips were placed into 6-well dishes seeded with 1 x 106 cells. Transfections were performed with FH-enhanced GFP fusion construct using 1 µg of DNA and 3 µl of FuGENE 6 transfection reagent (Roche Diagnostics, Laval, Canada), according to the manufacturer’s protocol. A pEGFPN3 empty vector was also transfected into CHO-K1 as a negative control. Twenty-four hours after transfections, the coverslips were rinsed with PBS, and cells were fixed with paraformaldehyde 4%. Mitotracker Orange CMTMRos (Invitrogen, Carlsbad, CA, USA) was employed as a control for the mitochondrial pattern. Slides were observed under a Leica fluorescence microscope and images were analyzed using Adobe Photoshop 7.0.
Generation of fly transformants
P-vector transformants were generated by standard embryo-injection methods, according to Rubin and Spradling (32)
. A total of 10 independent transforming lines were generated. The UAS-GAL4 system (33)
was carried out to generate the fh-RNAi and the fh-overexpressed flies. Every experiment was always done simultaneously for control (yw x GAL4 driver) and for fh responder flies (UAS-fhIR x GAL4 driver and UAS-fh x GAL4 driver) at 25°C or 29°C.
Quantitative real-time PCR
Total RNA was isolated from 100 males or from age-matched embryos using TriPure Isolation Reagent, according to the manufacturer’s instructions (Roche Diagnostics). cDNA was synthesized with Expand Reverse transcriptase (Roche Diagnostics) and oligo-dT primers. Quantitative real-time PCR was performed with ABI PRISM 7500 sequence detection system (Applied Biosystems, Foster City, CA). TaqMan probes for RP-49 (control) and fh containing 6-carboxyfluorescein (6-FAM) at the 5' end, and the appropriate primers were synthesized by Applied Biosystems. Data analysis was performed in triplicate experiments. Statistical significance was evaluated using Student’s t test and P < 0.05 was considered significant.
Western blot analysis
Total protein was extracted from age-matched embryos following the method previously described (34)
. Protein levels were quantified by Bradford assay. Fifty micrograms of total protein was applied to each lane. Samples were separated on 5% stacking, 15% resolving Tris-glycine SDS-polyacrylamide gels. Resolved proteins were electroblotted to Hybond enhanced chemiluminescence (ECL) nitrocellulose membrane (Amersham Biosciences) and probed with mouse anti-Frataxin monoclonal antibody (mAb) (MAB-10485, Immunological Sciences) in combination with goat anti-mouse IgG-horseradish peroxidase conjugate (Amersham Biosciences, Little Chalfont, Buckinghamshire, UK). Detection was carried out using ECL Detection Reagents (Amersham Biosciences). The mouse anti-
-tubulin was used in combination with goat anti-mouse IgG-horseradish peroxidase conjugate (Amersham Biosciences) as a control. The protein MW was estimated with a prestained protein molecular weight marker (Fermentas).
Immunocytochemistry staining
Whole mount embryos and stainings using the horseradish peroxidase technique were carried out as described previously (35)
. The primary antibodies used were: mouse anti-myosin heavy chain (anti-MHC) 1:8, mouse mAb 22C10 1:50, mouse 1D4 antifasciclin II 1:20, mouse mAb EC11 antipericardin 1:2, mouse mAb BP 102 anti-central nervous system (central nervous system (CNS)) axons 1:200 and rabbit Eve (Even-skipped protein) 1:1000.
Life span and climbing assays
Life span and climbing assays were performed as described in Botella et al. (36)
. Kaplan-Meier analysis of survival data with semiparametric log rank test was performed using the GraphPad Prism 2.0 software. Climbing data were examined for differences with a one-way ANOVA test, and means were compared using the SNK test. All statistical analyses were carried out with the package Statistical Packages for the Social Sciences (SPSS) V.12.0.1 and values of P < 0.05 were considered statistically significant.
Hyperoxia treatment
Hyperoxia treatment commenced 1 day posteclosion and was performed by exposing flies in a glass container with a constant flux of 99.5% oxygen under a low positive pressure at 25°C. Flies were confined in groups of 20 and were transferred every day to new vials containing regular food.
Mitochondria isolation
Mitochondria were isolated from 100 adult male flies after the procedure described in Bioxytech Aconitase-340TM with some modifications. Flies were homogenized in the appropriate buffer, and the resulting mashes were centrifuged twice at 800 g for 10 min at 4°C discarding the pellet every time. Supernatants were further centrifuged at 13000 g for 10 min at 4°C. Pellets were resuspended in 500 µl of ice-cold homogenization buffer at a concentration of 500 µg/ml. Mitochondrial pellets were disrupted by sonication for 30 s four times on a setting of 2 with 1 min. interval between each sonication.
For measurements of complexes I-IV respiratory rates, mitochondria were softly isolated to prevent break. Flies were placed in 500 µl of cold isolation buffer (154 mM KCl, 1 mM EDTA, pH 7.4), gently pounded and homogenates were filtered through 0.5 cm of cheesecloth placed in a new tube. Filtrate was centrifuged at 4°C for 8 min at 1500 g to remove cellular debris. Supernatants were discarded, and the resulting pellets were washed with 200 µl of isolation medium and finally resuspended in 50 µl of isolation medium. All subsequent assays were performed within 3 h, keeping the mitochondria suspensions on ice.
Enzyme assays
Aconitase activity was determined using Bioxytech Aconitase-340TM Spectrophotometric Assay kit. Succinate dehydrogenase was determined by the procedure described by Munujos et al. (37)
.
Mitochondrial respiratory assays
Rates of mitochondrial respiration (state 3, ADP-stimulated state and state 4, ADP-deleted state) were determined by oxygen consumption using a fiberoptic oxygen microsensor of Microx TX3 (Precision Sensing GmbH, Regensburg University, Germany). Temperature was maintained at 28°C and the reaction volume was 50 µl. Freshly isolated mitochondria were added to the respiration buffer (10 mM KH2PO4, 5 mM MgCl2, 120 mM KCl, 1.25 mg/ml BSA, pH 7.4) allowing to equilibrate for 30 s.
Complex I and complex II contribution was measured using NADH (60 mM) and succinate (10 mM) plus rotenone (100 µM), respectively. Pyruvate plus malate (200 µM each) were used to analyze the oxygen rate consumption after the activation of the Krebs cycle. Both substrates were added to the chamber and allowed to equilibrate for 30 s, followed by the addition of 5 mM ADP when required.
In the case of the respiration after addition of succinate plus rotenone (activation of complex II), only the state 3 values have been reported because no difference was observed with state 4, probably due to the inability of succinate to cross the inner mitochondrial membrane as has been previously observed by Ferguson et al. (38)
.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. To confirm the subcellular localization of FH, colocalization experiments using a frataxin-enhanced green fluorescent protein (FH-enhanced GFP) and a mitochondrial marker were performed. Transfection of a FH-enhanced GFP fusion construct into CHO-K1 cells revealed strong fluorescence signal in the mitochondrial reticulum that colocalized with the mitochondrion-selective probe mitotracker Orange CMTMRos (Fig. 1
). The subcellular localization of the Drosophila FH suggested an equivalent function to that of the human ortholog; therefore, a Drosophila strain showing a reduction of FH function would be of great value for modeling the progression of FA disease. In a model recently published by Anderson et al. (29)
, the general silencing of fh expression to undetectable levels resulted in lethality during development and, therefore, precluding any study of the effects of FH reduction in adult individuals.
|
Generation of RNAi frataxin flies
To bypass the preadult lethality obtained by the almost complete depletion of FH, we reduced fh expression to a level that allowed study of adult individuals. In this way, we generated a model that would more closely parallel the situation in FA patients.
Several Drosophila transgenic lines carrying the UAS-fhIR construct were generated. The presence of this construct was verified by Southern blot and after exclusion of position-dependent effects, a line carrying the UAS-fhIR construct on the second chromosome was selected for further analysis.
To test the efficiency of our model and reveal its potential, knockdown of FH was performed throughout the organism, as well as in specific tissues because most patients are affected by progressive polyneuropathy, myopathy and hypertrophic cardiomyopathy (39)
. FH knockdowns were generated using induction of RNAi by means of the GAL4/UAS system at 29°C. Table 1
shows the GAL4 lines used and the results obtained. The gene silencing performed using the ubiquitous drivers, actin-GAL4 and da-GAL4, resulted in lethality showing, in agreement with Anderson et al. (29)
, that the function of FH is essential during development. The lethality observed with 24B and Dot drivers, demonstrated that FH plays a key role in the development of the early mesoderm (precursor tissue of muscles and dorsal vessel) and heart. Silencing of fh using different nervous system drivers resulted in viable progeny without any gross morphological abnormalities. The offspring from these crosses were examined both for their life span and climbing ability. A statistically significant reduction of the life span (Fig. 2
A) and a 15% decline in climbing capability was found only in the case of neuralized-GAL4, which is expressed in all sensory organs and their precursors. The general silencing of fh in motoneurons (D42-GAL4), brain (c698a-GAL4), and dopaminergic neurons (Ddc-GAL4) did not result in any distinguishable phenotype (Fig. 2B-D
). The results above were in agreement with the clinical aspects described for FA patients, where sensory neurons seem to be the most affected component of the nervous system and validated the use of our RNAi model to further study FA pathology in flies.
|
|
Moderate reduction of fh expression shortens life span and impairs climbing ability
Next, we determined experimental conditions that could allow us to parallel more closely the situation in FA patients. Two parameters were crucial to achieve this: first, reduction of fh expression ubiquitously and second, presence of a given fh level able to circumvent developmental defects leading to lethality. We achieved the optimal level of fh silencing using the ubiquitous driver actin-GAL4 at 25°C. Under these conditions, the fh mRNA level detected using real-time PCR showed the presence of only 30% of fh messenger when compared to control flies. This amount of fh mRNA is similar to the expression levels reported in FA patients (7
, 40)
.
Survival analysis revealed a reduction of mean and maximum life span of 60% and 32%, respectively in fh-RNAi adult flies when compared to controls (Fig. 3
A). Moreover, we also found a reduction in climbing ability of 45% in 5-days-old fh-RNAi individuals (Fig. 3B
). These results suggested that the function of FH is not only essential for development but also required for the maintenance of vital functions of the adult individual. Alternatively, the reduction in survivorship and the decay in climbing performance may also be influenced by, in our hands, undetectable developmental defects.
|
fh mediates protection against oxidative stress
In Drosophila, reductions of life span and decline in climbing capabilities have been often related to increased sensitivity to oxidative damage (36
, 41)
. To test whether the phenotypes caused by reducing fh expression might be a consequence of an enhanced susceptibility to oxidative stress, we exposed fh-RNAi and control flies to a high oxidative atmosphere (99.5% O2). Hyperoxia has been found to be a relevant strategy to recreate the effect of unbalanced cellular oxidative status in Drosophila (42)
. Under these conditions, control flies showed an average maximum life span of 9 days, whereas fh-RNAi flies showed a 44% reduction in this value (Fig. 3C
). This result indicated that FH function might be involved in a protection mechanism against oxidative stress and suggested that oxidative damage plays a role in the phenotypes induced in fh-RNAi flies during the aging process. The association found between lack of FH function and oxidative stress is in agreement with findings reported in other models (22
23
24
25
26
27)
yet contradicts recent suggestions (28
, 29)
that there is no role for oxidative stress in FA pathology.
fh protects against oxidative stress-induced aconitase inactivation
Since the mitochondrial enzyme aconitase has been identified as a specific target of oxidative stress in a variety of organisms (43
44
45)
, and its activity is found to be seriously affected in FA tissues, we next assessed the functional integrity of this enzyme in our Drosophila model. As shown in Fig. 4
A, aconitase activity in control flies was reduced, as expected, during aging (46)
. Surprisingly, this reduction was similar in fh-RNAi flies, indicating that a decrease in the FH function did not appear to have an effect on the overall aconitase activity measured during the aging process (see Discussion). Interestingly, when aconitase activity was measured after one day of hyperoxia treatment, a dramatic reduction in the enzymatic activity was observed in flies defective for FH (Fig. 4B
). As reported in Das et al. (44)
, aconitase activity in wild-type (WT) flies decreases under hyperoxia, demonstrating the vulnerability of this enzyme to high levels of reactive oxygen species (ROS). As this reduction was enhanced 30% in fh-RNAi flies, the result shows the importance of FH function in protecting aconitase against its oxidative stress-induced inactivation. This result is in agreement with the proposed function for frataxin acting as an aconitase chaperone (17)
.
|
In contrast, the activity of succinate dehydrogenase (SDH), a mitochondrial Fe-S cluster containing enzyme also found affected in FA patients, showed no change in aging (Fig. 4C
) or after one day under hyperoxia conditions in controls and fh-RNAi flies (Fig. 4D
). Our results were in agreement with other reports that show that SDH is not especially sensitive to ROS during the normal aging process (47)
. Moreover, this fact could explain why in hyperoxic conditions, the activity of this enzyme does not decrease due to a deficit of FH.
Taken together, these results suggested that a decline in aconitase activity by ROS inactivation is one of the primary events responsible for the physiological defects observed in fh-RNAi flies.
Frataxin deficiency seriously compromises mitochondrial respiration
A decrease in aconitase activity would be predicted to impair the citric acid cycle, diminishing the supply of reducing equivalents for electron transport and, therefore, decreasing ATP production. Moreover, inhibition of aconitase activity has been shown to decrease oxygen consumption in cultured cells (48)
. To test whether the reduction of FH function would affect mitochondrial respiration, we measured oxygen consumption rates of isolated mitochondria from control and fh-RNAi flies in normoxia and hyperoxia conditions. No difference was found in state 3 (active respiration state) or state 4 (basal respiration state) between 1-day fh-RNAi and control flies under normoxic conditions (Fig. 5
A). A 37% reduction in oxygen consumption in state 3 was readily observed in controls when maintained 1 day under hyperoxia. Strikingly, a 73% decrease was found (state 3) in fh-RNAi flies after oxidative stress treatment, a two fold reduction when compared to the hyperoxic values in control flies (Fig. 5A
). This result indicated that on oxidative insult, the function of FH is necessary but not sufficient to warrant the supply of energy for the cells.
|
There can be several causes for the decreased respiration observed in FH-deficient flies. These include defects in respiratory complexes and/or a diminution of reducing equivalents necessary for electron transport from the citric acid cycle due to the aconitase deficiency described above. To distinguish between these two possibilities, oxygen consumption induced by stimulation of mitochondrial respiratory complex I and II was analyzed.
On addition of NADH (complex I), no difference was observed in oxygen consumption rate between control and fh-RNAi flies under normoxic or hyperoxic treatment, although hyperoxia was able to equally decrease this rate by 38% in both cases (Fig. 5B
). The analysis of respiration using succinate plus rotenone (complex II) showed no difference between controls and fh-RNAi flies under normoxia or hyperoxia treatment (Fig. 5C
). This was in agreement with the above-mentioned lack of effect on SDH activity observed in fh-RNAi flies. Altogether, these results indicated that complexes I and II (and indirectly III and IV) do not seem to be affected by the lack of FH under our experimental conditions. Thus, in our frataxin model, the oxidative inactivation of aconitase, and not the instability of the respiratory chain complexes, appears to be the initial cause of the observed reduction in respiration rate.
Overexpression of fh induces similar defects to its reduction
To study the consequences of increasing fh expression, several Drosophila transgenic lines carrying the UAS-fh construct were also generated. After exclusion of position-dependent expression of the transgene, a line carrying that construct on the second chromosome was selected for further analysis.
Overexpression analyses were performed at 29°C using the same GAL4 lines (ubiquitous pattern, nervous system, muscles and heart) as in the fh-RNAi experiments. The results obtained are summarized in Table 1
. An early and ubiquitous overexpression of fh with the drivers actin and da, led to the death of all individuals at early pupae stage and as third instar larvae, respectively. Full lethality was also observed with 24B and Dot drivers, similar to that seen in the fh-RNAi experiments. Although fh overexpression with Dot-GAL4 began during embryogenesis (49)
, 95% of individuals died in pupal stages with the remaining 5% dying during eclosion from puparium. The postembryonic lethality can be explained because Drosophila can complete embryonic development despite an affected heart, but the presence of a functional organ is required for the rest of the life cycle (50)
. Taking the RNAi and overexpression data together, these results suggested that the same process is being disrupted by either a FH excess or deficit and support a critical function of FH in the early development of muscles and heart. In addition, overexpression of fh using neural specific GAL4 drivers was compatible with normal preadult development. Thus, life span and climbing experiments were conducted to monitor age-related changes. Overexpression using neuralized (sensory organs) and D42 (motoneurons) GAL4 drivers resulted in a reduction of 64% and 49% of mean life span values (Fig. 2A, B
) and a 85% and 60% decline in climbing ability respectively, effects stronger than those dectected in the fh-RNAi flies. No phenotype was observed either with c698a (brain) or with Ddc (dopaminergic cells) GAL4 drivers (Fig. 2C, D
). These results reinforced the important role of FH activity in the peripheral nervous system (PNS) of Drosophila.
Ubiquitous overexpression of fh impairs development
To quantify the level of fh overexpression, fh-mRNA was detected in da-GAL4>UAS-fh embryos using real-time PCR. A 9-fold increase was found in these embryos when compared to controls. This induction of expression correlated with the results obtained in Western blot analysis (Fig. 6
A).
|
Immunohistochemistry was carried out to determine the underlying defects associated with the lethal phenotypes obtained from da-driven overexpression of fh. Anti-MHC staining revealed muscles clearly disrupted when compared to controls (Fig. 6B, C
). In almost all embryos, we unequivocally observed three specific partially formed muscles: dorsal acute 3, dorsal oblique 3, and dorsal oblique 4. To investigate whether the nervous system was also affected, two different neural markers (1D4 antifasciclin II and 22C10) were used. ID4 labels the 6 longitudinal axons of the CNS and the intersegmental (ISN) and segmental (SN) nerves; and 22C10 labels the PNS. We detected aberrant axonal tracks in 70% of da-GAL4>UAS-fh embryos when using the 1D4 staining (Fig. 6D, E
). 22C10 staining showed an increase in the number of sensory ventral neurons (Fig. 6F, G
) and axonal pathfinding defects (Fig. 6H
) in 10% of the embryos. In contrast, no abnormalities were detected in CNS with the Eve and BP 102 stainings (data not shown).
We also analyzed the possible defects caused by fh overexpression in the heart, which is a key affected tissue in FA. For this purpose, Dot-GAL4>UAS-fh embryos were stained with ECII antibody (Ab), which labels the extracellular matrix (ECM) surrounding the pericardial and cardial cells of the heart tube. This staining revealed the lack of some pericardial cells along the tubular structure of the developing heart (Fig. 6I, J
).
Taken altogether, these results indicated that overexpression of FH seriously affects the development of embryonic muscles, peripheral nervous system, and the heart. We find that the key tissues affected in FA patients are unexpectedly sensitive to increase of frataxin function in Drosophila.
Frataxin overexpression inhibits mitochondrial aconitase activity under hyperoxia
To check whether oxidative stress might be a factor involved in the overexpression phenotypes, we measured the activity of mitochondrial aconitase in actin-GAL4>UAS-fh adult flies at 25°C (at this temperature, overexpression of fh using the ubiquitous driver actin-GAL4 provided adult individuals). When total aconitase activity was determined during aging, no differences were detected between controls and overexpressed flies (data not shown). However, when flies were exposed to a high oxidative atmosphere (99.5% O2), fh-overexpressed flies showed a 40% reduction in aconitase activity compared to control flies (Fig. 4B
). This result reinforced the relationship between FH and aconitase in Drosophila and the role of the oxidative stress in the phenotypes induced by fh misexpression.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In this work, we have confirmed the mitochondrial localization of FH, suggesting that the molecular processes relevant to the disease might also be shared in a Drosophila model. Therefore, we report a new RNAi model of FA in Drosophila, in which we developed experimental conditions producing a moderate systemic reduction of frataxin level, which, in both humans and flies, is compatible with normal embryonic development but results in clinical symptoms associated with FA. We also report the effects of an fh overexpression system, which provides additional insight to frataxin function.
In agreement with the essential role for frataxin during development proposed in previous animal models (29
, 51)
, we have found that the silencing of fh using general and some tissue specific GAL4 lines at 29°C also resulted in lethality. Furthermore, we found a distinct sensitivity of different tissues to the reduction of FH. Silencing of fh in muscle and heart was lethal, whereas reduction of fh expression in neural tissues did not result in preadult phenotypes. This shows that FH function in muscles and heart is critical during embryonic development, whereas FH activity in the nervous system appears necessary during adulthood and is restricted to the PNS.
In this work, we have created a scenario that results in a three-fold reduction of fh mRNA levels by combining our RNAi transgenic flies with the actin-GAL4 driver at 25°C. This reduction was compatible with a normal embryonic development but resulted in shortened life span and reduced climbing ability in adulthood. These results strikingly parallel the situation in FA patients, whose motor abilities are mainly impaired from puberty onward and show a decreased life expectancy (52)
.
The link between oxidative stress and FA is controversial. Variation in levels of oxidative stress markers in FA patients have been used to monitor the development of the pathology (19
20
21)
, and a relationship between oxidative stress and progression of the FA disease has been already proposed (18
, 53)
. Recently, two reports have dismissed the role of oxidative damage in the progression of the disease (28
, 29)
. To provide insight into this controversy, we investigated the role of the oxidative insult in the development of the pathology using our model system. One day actin-GAL4/UAS-fhIR individuals were exposed to hyperoxia. The fh-RNAi flies showed a hypersensitive response to hyperoxia, thereby supporting a causative role of oxidative stress in FA. This result also suggested that the shortened life span and locomotor defects observed during aging in fh-RNAi flies is due to the deleterious effects of ROS during their lifetime.
Since mitochondrial aconitase is the most affected enzyme in FA and it has been shown to be, in contrast to other tricarboxylic acid (TCA) cycle enzymes, selectively inactivated by oxidative stress (44
45
46)
, we monitored its activity in fh-RNAi flies. Interestingly, the exposure to hyperoxia led to an extreme reduction of aconitase activity in the fh-RNAi flies. This result indicated a role for FH in protecting aconitase activity against ROS-mediated inactivation, supporting frataxin function as an aconitase chaperone (17)
. Oxidative stress induced the release of the Fe- aconitase Fe-S cluster leading to its inactivation (54)
. Thus, without the protection of frataxin, aconitase might not be reactivated at normal rates when oxidative stress reaches high levels, as happens in hyperoxia or during the aging process, resulting in its irreversible inactivation.
Surprisingly, during aging, no difference between control and fh-RNAi flies was detected when total aconitase activity was determined, though the activity showed as expected an age-dependent reduction. We have shown in this work that the reduction of FH function distinctively affects diverse tissues. In addition, the different sensitivity of some tissues to oxidative stress might account for a dissimilar rate of aconitase inactivation; therefore, the lack of a distinct decrease of total aconitase activity in fh-RNAi flies could be due to the masking effect of intact aconitase when whole fly extracts were used.
Succinate dehydrogenase activity did not show significant alterations during aging or hyperoxia in control or silenced flies. Altogether, these results suggested that, at least in our model, the decay in aconitase activity is the primary event in the development of the pathology, occurring before the reduction of SDH activity reported in patient tissues and in other animal models (23
, 55
, 56)
.
Strikingly, overexpression of FH also enhanced the oxidative-mediated inactivation of mitochondrial aconitase. It suggests that an excess of frataxin function is inducing defects in Drosophila likely in an aconitase-related manner as occurs when frataxin expression is diminished. The inactivation of this enzyme would probably compromise the TCA cycle and the energetic supply from oxidative phosphorylation leading to the lethality observed. This may also explain why tissues sensitive to FH reduction are sensitive to FH increase.
Aconitase synthesis requires an intact [4Fe-4S] cluster for its function and may be affected in an environment with limited iron availability. Overexpression of mitochondrial ferritin, a protein related to iron metabolism in the cell, has been reported to inhibit the activity of the mitochondrial aconitase (57)
. Frataxin has been shown to form in vitro iron-binding multimers (10)
, whose structure would be similar to ferritin macromolecules. Therefore, overexpression might induce a new in vivo function for frataxin, being partially responsible of the Fe nonavailability in the mitochondria conducting to aconitase dysfunction. Alternatively, physical interactions between frataxin and other proteins such as SDH, aconitase, ferrochelatase, or Isu1 have been reported in different organisms (14
, 17
, 58
, 59)
. Since those genes are also present in Drosophila (60)
and our results support a direct interaction with one of them, all of these interactions may also exist in the fly. An excess of frataxin might saturate that set of interactions leading to dysfunction of the TCA cycle, the respiratory chain and/or the FeS cluster assembly machinery, provoking an energetic catastrophe. This interpretation suggests that overexpression might be acting as a dominant negative mutation.
TCA cycle plays a pivotal role in mitochondria bioenergetics, and because of its biochemical design, a decrease in the activity of one enzyme will potentially compromises the entire respiratory process. The decline in aconitase activity could well explain the reduction in ATP production observed in FA patients after exercise (61
, 62)
. To test that possibility, the oxygen consumption rate of isolated mitochondria in our model was measured. Our results showed that hyperoxia led to a remarkable reduction in oxygen consumption rates in mitochondrial extracts in our fh-RNAi flies. The fact that this reduction was not detected by activating specifically complex I or II and only found on addition of pyruvate plus malate indicates that, in fh-RNAi flies, the production of electron supply from the TCA cycle is defective.
The lack of respiration defects found by specifically stimulating complex II is also in agreement with the intact SDH activity determined under the same conditions and suggests mitochondrial aconitase as the primary target of FH deficiency in the silenced flies.
According to our results, we suggest that in Friedreich ataxia, the regular oxidative mediated inactivation of aconitase occurring normally during the aging process is enhanced due to the lack of frataxin. When mitochondrial aconitase function is seriously compromised, a failure of the TCA cycle leads to an energetic catastrophe (63)
. This could explain why in FA patients, tissues with a high energetic demand such as skeletal muscle and heart are seriously affected (39
, 61
, 64)
. Aconitase inactivation may initiate a harmful feed back loop by increasing oxidative stress in two ways: 1) an impaired electron transport chain would induce the accumulation of substrates such as NADH that by autooxidation might generate a "reductive stress" condition increasing the production of ROS (43)
and second, leading to an increase in citrate concentration, the immediate substrate of aconitase. Such accumulation has been already observed in mammals and correlates with the drop of aconitase activity during aging (65)
. Citrate is a known iron chelator and its toxicity is iron dependent in a yeast frataxin-deficient strain (66)
. Moreover, citrate Fe3+ complexes have been shown to increase oxidant damage in vitro (67)
. Interestingly, cardiac tissue shows the highest concentration of citrate (68)
, and in FA patients, it is where iron deposits have been found (39)
. This burst in oxidative damage might accelerate aconitase inactivation and affects other macromolecules, such as the Fe-S enzymes of the respiratory chain.
We have shown in a multicellular model organism that frataxin is primarily involved in protecting aconitase against oxidative-mediated inactivation and propose that oxidative stress plays a central role in the progression of Friedreich ataxia.
|
ACKNOWLEDGMENTS |
|---|
We thank J. C. Adell for his advice on the statistical analyses and the Servicio Central de Soporte a la Investigación Experimental de la Universitat de València for access to DNA analysis resources and databases.
|
FOOTNOTES |
|---|
Received for publication January 23, 2006. Accepted for publication September 14, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
21 kDa and the mature form of