| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online March 20, 2001 as doi:10.1096/fj.00-0755fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
University of South Alabama College of Medicine, Departments of Pharmacology and Cell Biology and
* Neuroscience, Mobile, Alabama 36688, USA
2Correspondence: Department of Pharmacology, College of Medicine, MSB 3130, University of South Alabama, Mobile, AL 36688, USA. E-mail: mgillesp{at}jaguar1.usouthal.edu
SPECIFIC AIMS
The conclusion that mitochondrially derived oxidants play a role in hypoxic signal transduction has remained controversial in part because of doubts concerning methods of free radical detection and because of theoretical objections to the notion that mitochondria are capable of increasing oxidant generation in a low oxygen environment. To help resolve this controversy, the present study tested the hypothesis that hypoxia causes oxidative modifications in mtDNA or in the nuclear vascular endothelial cell growth factor (VEGF) gene, the latter of which should serve as a negative control based on the relatively insensitive nuclear genome to oxidant stress in comparison to that in the mitochondria.
PRINCIPAL FINDINGS
1. Hypoxia causes an oxidant stress in pulmonary artery endothelial
cells as detected by fluorescence of dichlorfluorescein (DCF)
We used rat cultured main pulmonary artery cells (PAECs) as the
model system for these studies because they are important in the
integrated pulmonary vascular response to hypoxia and because hypoxia
is known to promote multiple direct effects on PAECs in culture. PAECs
cultured on petri plates were mounted on a temperature-controlled stage
of a laser confocal microscope and incubated with the nonfluorescent
2'-7'-dichlorodihydrofluorescein diacetate. The diacetate moiety is
cleaved after cellular uptake to dichlorodihydrofluorescein, which, in
the presence of reactive oxygen and nitrogen species, is oxidized to
the fluorescent dichlorfluorescein. After a normoxic stabilization
period, PAECs were exposed to a hypoxic environment by perfusing the
airspace above the cells with a mixture of 95% nitrogen and 5%
CO2. Within 10 min of initiating the hypoxic
challenge, DCF fluorescence was approximately 40% above baseline; by
40 min, fluorescence intensity exceeded baseline by nearly 200%. These
observations are consistent with findings in a number of other cells
and suggest that hypoxia promotes an oxidant stress in PAECs.
2. Hypoxia causes oxidative modifications in the nuclear VEGF gene
promoter that are prominent in the hypoxic response element
Quantitative Southern blot analysis was used to determine whether
hypoxia promoted oxidative modifications in the mitochondrial genome
and in a 5 kb sequence of the nuclear VEGF gene 5' promoter region.
Counter to expectations, integrity of mtDNA was unaffected by hypoxia.
However, as shown in Fig. 1
, assessment of alkali-treated DNA to detect oxidative modifications in
the deoxyribose backbone revealed discernible decreases in
hybridization intensities at early time points after hypoxic exposure.
Treatment of DNA with endonuclease III, a DNA repair enzyme specific
for the recognition and cleavage of the DNA strand at sites of
oxidative modifications to pyrimidine bases, was associated with
further decreases in hybridization intensity. When data from four such
experiments were pooled and lesion density calculated according to an
equation well entrenched in the literature, there were impressive
increases in alkali and Endo III-detectable modifications as early as
3 h of hypoxic exposure, which peaked at 6 h and then
declined to near control levels within 48 h despite the
persistence of hypoxic exposure.
|
3. Hypoxia-induced base modifications in the VEGF promoter are
prominent in the hypoxic response element
To determine the nucleotide specificity of the hypoxia-induced
oxidative modifications in the nuclear VEGF gene, ligation-mediated PCR
(LMPCR) analysis was performed on a 50 bp sequence of the VEGF promoter
containing the HIF-1 and AP-1 response elements. A spatial map of
lesions was constructed by plotting a bar, the height of which
reflected the prevalence of damage, as a function of the position of
each nucleotide in the sequence of interest. This analysis was applied
to four LMPCR experiments in which DNA was treated with alkali alone
and three where DNA was treated with alkali plus endonuclease III. The
lesion map shown in Fig. 2
displays hypoxia-induced, alkali-detectable lesions in an 
50
sequence of the VEGF promoter. The modifications could be characterized
as numerous, widely scattered, and present in both the consensus AP-1
and HIF-1 response elements. A cytosine in a 5' palindromic CACA
sequence flanking the HIF-1 site, believed to be required for
transactivation of the response element, also was oxidatively modified.
Endonuclease III treatment intensified the prevalence of lesions almost
exclusively at pyrimidines (data not shown).
|
5. Hypoxia-induced oxidative modifications in the VEGF promoter are
different from those evoked by reactive species generated by exogenous
xanthine oxidase (XO)
Using LMPCR analyses, we compared the pattern of lesions evoked by
XO in the VEGF promoter to that induced by hypoxia. In preliminary
studies, we confirmed previous findings that mtDNA was more sensitive
to XO than the nuclear VEGF gene. In this regard, quantitative Southern
blot analysis failed to detect lesions in the VEGF promoter at XO doses
up to 50 mU/kg, which caused complete degradation of the mitochondrial
genome. However, using a 1 h treatment with 400 mU/ml XO to
overcome the inherent insensitivity of the nuclear genome to oxidant
stress, we were able to detect relatively low-intensity lesions in the
VEGF promoter by LMPCR. The lesion map for XO-induced modifications in
the VEGF promoter, also displayed in Fig. 2
, indicates that XO-mediated
damage is less prevalent than hypoxia and exhibited a different
nucleotide specificity. Neither the HIF-1 nor AP-1 response elements
were damaged as prominently by XO as they were in hypoxia. As an
additional means of comparing the effects of hypoxia and XO, we
determined the proportion of each nucleotide modified by the two
stimuli expressed in terms of the total number of that nucleotide
within the sequence of interest. In the coding strand, guanines,
adenines, and cytosines were more frequently damaged by hypoxia in
comparison to XO, whereas in the noncoding strand, hypoxia
preferentially damaged guanines and adenines.
CONCLUSIONS AND SIGNIFICANCE
Contrary to our expectations, hypoxia failed to erode the integrity of the mitochondrial genome; instead, as determined by both endonuclease fingerprint and LMPCR analyses, it caused prominent oxidative lesions in the VEGF promoter. DNA damage revealed by alkali treatment is most likely due to oxidant attack on the deoxyribose backbone leading to single strand breaks or formation of abasic sites. Additional support for an oxidant-dependent pathway of hypoxia-induced DNA modifications derives from the ability of endonuclease III to unmask a considerable increase in lesion density. Endonuclease III, a well-characterized bacterial DNA repair enzyme, specifically recognizes pyrimidine oxidation products and cleaves the DNA strand at such sites.
The most surprising and potentially significant observations in the present study relate to the distribution, nature, and sequence specificity of hypoxia-induced oxidative DNA modifications. 1) Unlike toxic oxygen radicals (e.g., XO, menadione, and alloxan), which predominantly damage the mitochondrial genome, hypoxia-induced modifications were prominent only in the nucleus. This general pattern of lesion formation has not, to our knowledge, been previously observed for any other oxidant-generating stimulus. 2) Whereas in other cells the damage evoked by toxic oxygen radical generators is characterized by alkali-detectable lesions to the deoxyribose backbone, the endonuclease III fingerprint associated with hypoxic exposure is indicative of oxidative modifications to pyrimidine bases. The chemical nature of hypoxia-induced DNA modifications thus appears to differ from that associated with toxic oxygen radical-generating stimuli. 3) A third remarkable feature of the hypoxia-induced modifications to the VEGF promoter is the prevalence of lesion sites within the hypoxic response element. Specifically, our LMPCR analysis revealed widespread lesion formation with ‘hot spots’ prominent at critical nucleotides in the consensus AP-1 and HIF-1 DNA binding sequences, both of which have been linked to hypoxic gene regulation in a variety of cell types.
The above considerations point to an over-riding question: What is the
biological significance of these unusual oxidative base modifications
that occur in an important regulatory region of a gene during
adaptation to a hypoxic environment and in the absence of overt
cytotoxicity? At one extreme, it is possible that the lesions have no
role in the adaptive response. In view of the multiple mechanisms used
by cells to detect, repair, and respond to oxidative lesions, this
seems unlikely. As displayed in Fig. 3
, there are at least several mechanisms whereby hypoxia-induced
modifications in nuclear genes could be significant in terms of
adaptation to the hypoxic environment or cellular aging. First, DNA
modifications could serve as a trigger for critical cell cycle and cell
survival genes that coordinate the specific adaptive response to the
hypoxic environment with cell cycle traversion. In this regard, mRNA
transcripts for members of the growth arrest and DNA damage family of
stress proteins, whose expression is directly induced by DNA damage,
are increased by hypoxia in both transformed and bovine pulmonary
artery smooth muscle cells. In nonvascular transformed cells, the
regulator of apoptosis and other vital cell functions, p53, is elevated
by culture in hypoxic conditions. Like the GADD genes and others, p53
can be directly induced by DNA damage. Second, it should be considered
that the formation of oxidized bases within the hypoxic response
element could modify local sequence topography and thereby affect
binding of transactivating factors. In this manner, the hypoxia-induced
base oxidation lesions could play a dynamic regulatory role in
governing gene expression. Finally, it is possible that DNA lesions
associated with hypoxic signaling could contribute to cellular
senescence. We noted that the equilibrium lesion density in the VEGF
promoter had returned to control levels within 48 h despite
continuation of the hypoxic environment, implying diminished production
of DNA damaging free radicals and/or the operation of DNA repair
pathways. If lesions were to persist with advancing age because of
defective repair, for example, the process of senescence could be
accelerated. These hypotheses, as well as others that could be
formulated on the basis of the current observations, will require
additional investigation.
|
In summary, the present study shows in PAECs that hypoxia causes oxidative lesions in nuclear, but not mitochondrial, DNA. This observation provides additional evidence, not obscured by the selectivity and specificity issues surrounding other approaches to detect oxidant production and effects, that hypoxia does promote oxygen radical generation that modifies an important macromolecular target. In addition, the surprising finding that nuclear DNA, including functionally significant sequences within the VEGF promoter, is oxidatively modified in hypoxia raises new questions about the mechanism(s) by which reactive species generated in the context of physiologic and pathologic signaling govern cellular responses.
FOOTNOTES
1 To read the full text of this article, go to
http://www.fasebj.org/cgi/doi/10.1096/fj.00-0755fje ; to cite this
article, use FASEB J. (March 20, 2001)
10.1096/fj.00-0755fje 
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
