www.elsevier.com/locate/yexnrExperimental Neurology 189 (2004) 150–161

Transient exposure of rat pups to hyperoxia at normobaric and hyperbaric

pressures does not cause retinopathy of prematurity

John W. Calvert,a,b Changman Zhou,a and John H. Zhanga,b,*

aDepartment of Neurosurgery, Louisiana State University Health Sciences Center, Shreveport, LA 71130-3932, USAbDepartment of Molecular and Cellular Physiology, Louisiana State University Health Sciences Center, Shreveport, LA, USA

Received 5 March 2004; revised 23 April 2004; accepted 18 May 2004

Available online 3 July 2004

Abstract

We have shown that hyperoxia reduces brain damage in a rat model of hypoxia– ischemia. The purpose of this study was to examine the

possibility of hyperoxia in inducing vision-threatening retinopathy. Two different experiments were conducted in this study. PART 1: seven-

day-old rat pups were subjected to unilateral carotid artery ligation followed by 2 h of hypoxia (8% O2 at 37jC). Pups were treated with

100% oxygen at 1 ATA, 1.5 ATA, and 3.0 ATA for a duration of 1 h. PART 2: Newborn rat pups were exposed to 100% oxygen at 1, 1.5, or

3.0 ATA for 1 h, the same treatment protocol used for brain protection after hypoxia– ischemia. Retinopathy was evaluated by the degree of

neovascularization (measuring retinal vascular density), by the structural abnormalities (histology) in the retina, and by the expression of

hypoxia–hyperoxia sensitive proteins including hypoxia-inducible factor-1a (HIF-1a) and vascular endothelial growth factor (VEGF) at 24

h, 1, 2, and 10 weeks after hyperoxia exposure. Hyperoxic treatment at all pressures administered significantly reduced the hypoxia–

ischemic-induced reduction in brain weight. Retinal vascular density measurements revealed no signs of neovascularization after hyperoxia

exposure. There were also no abnormalities in the structure of the retina and no changes in the protein expression of HIF-1a and VEGF

following hyperoxia exposure. Exposure to hyperoxia for 1 h at normobaric or hyperbaric pressures did not result in the structural changes

or abnormal vascularization that is associated with retinopathy of prematurity, suggesting that hyperoxia is a safe treatment for hypoxic

newborn infants.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Retinopathy; VEGF; HIF-1a; HBO; Neonates

Introduction

The emergence of retinopathy of prematurity as a leading

cause of blindness in infants has occurred over the past 60

years, seemingly as a result of the advances in neonatal

intensive care practices that have allowed for the survival of

premature infants who have significant immature retinal

vasculature (Brooks et al., 2001; Miyamoto et al., 2002;

Smith, 2002; Stout and Stout, 2003). The incidence for

threshold (disease progression to the point of necessitating

peripheral retinal ablation therapy) retinopathy of prematu-

rity for premature infants weighing <1.25 kg is roughly 5%,

with about 20–30% of these infants becoming blind despite

treatment (Palmer, 2003; Reynolds, 2001).

0014-4886/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.expneurol.2004.05.030

* Corresponding author. Department of Neurosurgery, Louisiana State

University Health Sciences Center, 1501 Kings Highway, Shreveport, LA

71130-3932. Fax: +1-318-675-8805.

E-mail address: johnzhang3910@yahoo.com (J.H. Zhang).

Oxidant stress appears to play a role in the retinal vaso-

obliteration associated with retinopathy of prematurity

(Beauchamp et al., 2002; Weinberger et al., 2002). Exposure

to hyperoxia affects developing retinas by leading to micro-

vascular degeneration which produces inner retinal hypoxia,

which in turn leads to structural and functional changes.

These changes can lead to abnormal vascularization resulting

in the development of vision-threatening retinopathy (Beau-

champ et al., 2002; Brooks et al., 2001). Hypoxia–ischemia

is a common cause of brain injury in the perinatal period

leading to mental impairment, seizures, and permanent motor

deficits, such as cerebral palsy (Ferriero, 2001; Johnston,

2001). The role of supplemental oxygen therapy in the

development of retinopathy of prematurity was suggested

in the 1950s (Stout and Stout, 2003) and is a main reason for

not giving oxygen to hypoxic or premature infants (Neuba-

uer, 2002). But, recent data indicate that it is the withdrawal

from the oxygen environment that causes retinopathy of

prematurity and that subsequent oxygen exposure can cure

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161 151

the situation (The STOP-ROP Multicenter Study Group,

2000; Neubauer, 2002). We have shown that a single treat-

ment of hyperbaric oxygen (HBO) (100% oxygen) for 1 h at 3

atmospheres absolute (ATA) attenuates brain damage caused

by a hypoxia–ischemia insult on the neonatal rat brain by

reducing the progression of apoptotic neuronal injury and

increasing sensorimotor function (Calvert et al., 2002, 2003).

Consideration of HBO treatment for hypoxia–ischemia is

derived from the belief that oxygen under increased pressure

might salvage the still viable, though nonfunctioning, tissue

surrounding the area of insult by increasing the amount of

oxygen dissolved in blood plasma, thereby increasing oxygen

delivery to the brain (Nighoghossian and Trouillas, 1997).

Also, by raising the amount of oxygen available to the brain

tissue, hyperbaric oxygen may trigger a mechanism control-

ling cellular and vascular repair, presumably allowing time

for collateral circulation to develop (Mink and Dutka, 1995).

But, questions still remain about the safety of elevated levels

of oxygen for the treatment of newborns, because experi-

mentally, it has been shown that newborn rat pups exposed to

oxygen at 1.8 ATA for 10 days did not show any signs of

retinopathy of prematurity (Ricci and Calogero, 1988),

whereas when newborn rat pups were exposed to a higher

pressure of 5 ATA for 5 h, retinopathy of prematurity was

evident (Torbati et al., 1995). Therefore, the purpose of this

study was to test the hypothesis that a single exposure to

100% oxygen for 1 h at various pressures (1, 1.5, and 3.0

ATA) would not produce retinopathy of prematurity in

newborn rat pups.

Materials and methods

Groups and oxygen exposure

The Animal and Ethics Review Committee at the Loui-

siana State University Health Sciences Center-Shreveport

evaluated and approved the protocol used in this study.

Timed pregnant female Sprague–Dawley rats were obtained

from Harlan Labs (USA). After birth, pups were housed

with the dam under a 12:12 h light/dark cycle, with food and

water available ad libitum throughout the study. This study

was divided into two parts.

Part 1: the model used in this study is based on the Rice–

Vannucci (Rice et al., 1981) model previously described

(Calvert et al., 2002, 2003). Unsexed 7-day-old (day 0 =

day of birth), Sprague–Dawley (Harlan) rats were anesthe-

tized by inhalation with isoflurane (0.1%) in oxygen. The

pups were kept at a temperature of 37jC as the right common

carotid artery of each pup was exposed and ligated with 5–0

surgical sutures. The duration of the anesthesia did not exceed

20 min, and the pups were allowed to recover with their dams

for 2 h. They were then placed in a jar perfused with a

humidified gas mixture (8% oxygen balanced nitrogen) for 2

h. Both the jar and the gas mixture were kept at 37jC. Thepups were returned to their dams after the hypoxic exposure.

The pups that underwent hyperoxia treatment were allowed

to recover from the hypoxic exposure for 1 h before being

placed in the HBO chamber (Sechrist Industries, Inc., Ana-

heim, CA). The hyperoxic treatment of 100% oxygen was

administered at pressures of 3 atmospheres absolute (ATA),

1.5 ATA, and 1.0 ATA for 1 h and the pups were then returned

to their cages and dam after the treatment. Only one hyper-

oxic exposure was conducted for each pup. These pups were

sacrificed 2 weeks after the hypoxic–ischemic insult. The

pups in this part of the study were divided into the following

groups: (1) control group (n = 5), (2) hypoxia–ischemia (HI)

group (n = 15), (3). HI + 3.0 ATA group (n = 15), HI + 1.5

ATA group (n = 15), (4). HI + 1.0 ATA group (n = 15).

Part 2: unsexed 7-day-old (day 0 = day of birth) pups

were divided into the following groups: (1) control group

(n = 28), (2) normobaric group, 1.0 ATA group (n = 34), (3)

1.5 ATA group (n = 30), (4) 3.0 ATA group (n = 32), and (5)

retinopathy of prematurity group (n = 12). Pups were placed

in a hyperbaric oxygen chamber (Sechrist Industries, Inc) to

expose them to oxygen. For 1, 1.5, and 3 ATA groups,

oxygen (100%) was administered for duration of 1 h at the

following pressures: 1 (normobaric), 1.5, and 3.0 ATA. Each

pup only received one exposure. Control pups were not

exposed to oxygen. The pups were sacrificed at various time

points after hyperoxia exposure: 24 h, 1, 2, and 10 weeks.

For the retinopathy of prematurity group, postnatal day 5

(P5) pups were exposed to 100% oxygen at a pressure of 3

ATA for 10 h. These pups were sacrificed at 24 h and 2

weeks to obtain positive controls for retinopathy.

Brain weight

The pups from Part 1 were sacrificed under deep anes-

thesia. After removal of the brain, the cerebellum and brain

stem were removed from the forebrain. The hemispheres

were separated by a midline incision and then weighed on a

high precision balance (sensitivity F 0.001 g). The cerebel-

lum was also weighed. Brain damage was expressed as the

percent reduction of the ipsilateral (right) hemisphere com-

pared to the contralateral (left) hemisphere.

Retinal vascular density

Retinal vascular density was determined as previously

described (Torbati et al., 1995). The thorax was opened and

the left ventricle was perfused with 4 ml of India ink (Design

Higgins, Sanford, USA) for pups in the 1 and 2 weeks groups

and 12 ml for the rats in the 10-week group. After infusion of

ink, eyes were enucleated and placed in 2% formalin. Under a

dissecting microscope, the retinas were separated placed on

glass microscope slides and mounted with coverslips. Images

were captured using an imaging system that included an

Olympus BX51 microscope with MagnaFire capturing soft-

ware (SP2.1B) at magnifications of 1.25� and 10� and

examined with Scion Image for windows (Scion Corp.).

The retinal vascular density within an image was defined as

Fig. 1. Brain weight 2 weeks after hypoxia– ischemia and subsequent

hyperoxia exposure. #P<0.05 compared with control, **P < 0.05 compared

with HI, and yP < 0.05 compared with both control and HI (ANOVA).

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161152

the ratio of the total number of pixels representing the retinal

vessels divided by the total number of pixels representing

both the retinal tissue and retinal vessels.

Histology and cell counts

At the appropriate time (24 h, 2 and 10 weeks) after

hyperoxia exposure, rats (n = 3 per group) were anesthetized

with a-chloralose (40 mg/kg, i.p.)/urethane (400 mg/kg, i.p.).

The thorax was opened and the left ventricle was perfused

with 60 ml of 0.1M PBS (pH 7.4) followed by 60 ml of 2%

glutaraldehyde and 2% formalin in 0.1M PBS (pH 7.4).

Following perfusion, the eyes were enucleated and post-fixed

in the same fixative and stored at 4jC. The eyes were then

placed in 3% sucrose just before being sectioned. The eyes

were cut across the conjunctiva and the cornea and crystalline

lens were removed. The eye was then cut in half and placed in

tissue freezing medium (Triangle Biomedical Sciences,

USA). Four-micron sections were cut using a cryostat (Leica

LM 3050S). Sections prepared from samples taken at 24

h after hyperoxia exposure were used for immunohistoche-

mistry and sections prepared from samples taken at 2 and 10

weeks were stained with hematoxylin and eosin.

To analyze the retinal sections stained with hematoxylin

and eosin, images of intact sections were captured at a

magnification of 40� and analyzed using Image-Pro Plus

4.5.1 software. For each eye, two slides containing four

sections were analyzed (representing whole eye). The num-

ber of neurons in the inner nuclear layer (INL), the number

of neurons in the outer nuclear layer (ONL), the thickness of

the outer plexiform layers (OPL), and the areas of both the

INL and ONL were determined. For each section, these

measurements were made in five distinct areas of the retina

(see schematic in Fig. 4). The counts from each of these

areas were then averaged to represent the counts for each

section. The counts for each of these sections were then

averaged to represent the counts for each eye.

Immunohistochemistry

Immunohistochemistry was done as previously described

(Zhou et al., 2003) using an ABC Staining System (Santa

Cruz Biotechnology, USA). Briefly, sections prepared from

the samples taken at 24 h after hyperoxia exposure were

used for immunohistochemistry. These sections were

allowed to air dry before being washed in 0.01M PBS

followed by incubation in 3% hydrogen peroxide (H2O2)

to prevent reactions with endogenous peroxidases. This was

followed by washing in PBS and then incubation with 3%

normal serum in PBS (blocking solution). Sections were

then incubated with either rabbit anti-VEGF antibodies

(147) or rabbit anti-HIF-1a antibodies (H206) (Santa Cruz

Biotechnology) diluted in blocking solution (1:100) over-

night at 4jC followed by washing in PBS and incubation

with anti-rabbit secondary antibodies. After washing again

in PBS, sections were incubated with avidin–peroxidase

complex solution containing avidin–peroxidase conjugate.

The sections were then washed in PBS and peroxidase

activity was revealed by dipping the sections in a mixture

containing 3-3V diaminobenzidine (DAB) and H2O2 fol-

lowed by washing in distilled water. All the procedures were

conducted at room temperature. Sections were allowed to air

dry before being dehydrated and mounted with coverslips.

Images of the slides were captured using the imaging system

described above. Application of a control serum instead of

the primary antibody on sections provided negative controls.

Statistical analysis

The retinal vascular density values as well as the values

from the histological analysis are expressed as the means

F SEM. The values were analyzed by one-way analysis of

variances (ANOVA), followed by Tukey test. A P value less

than 5% was considered significant.

Results

Brain weight

To demonstrate that elevated levels of oxygen at normo-

baric and hyperbaric pressures can be neuroprotective, we

subjected neonatal rats to a hypoxic–ischemic insult fol-

lowed by 1 h of 100% oxygen at 1 ATA, 1.5 ATA, and 3

ATA. Two weeks after the insult and subsequent oxygen

treatment, the pups were sacrificed and the brains of those

pups were divided into two regions, contralateral hemi-

sphere and ipsilateral hemisphere. Brain damage was then

assessed by dividing the ipsilateral hemispheric weight by

the contralateral hemispheric weight and expressing this as a

percentage (Fig. 1). This rat model of hypoxia–ischemia

yields a reproducible pattern of hemispheric injury ipsila-

teral, but not contralateral to the ligated carotid artery (Han

et al., 2000). Since animals that are subjected to a hypoxic–

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161 153

ischemic insult show retardation in brain growth, the percent

reduction in ipsilateral brain weight to contralateral brain

weight allows for the evaluation and testing of neuropro-

tective agents and strategies. The ipsilateral hemisphere was

Fig. 2. Flat mounts of retinas of newborn rats exposed to hyperoxia at 1 atmosphe

week after hyperoxia exposure. (B) Inversion of retinal images in A to provide a b

Inversion of images in C. (E) Ten weeks after hyperoxia exposure. (F) Inversion

found to be 54.3% of the contralateral hemisphere after the

hypoxic– ischemic insult and hyperoxic treatment at all

pressures administered significantly reduced this damage,

as brain weights were found to be 78.168%, 81.230%, and

re absolute (ATA), 1.5 ATA, or 3.0 ATA and age-matched controls. (A) One

etter look at the vasculature. (C) Two weeks after hyperoxia exposure. (D)

of images in E. Scale bar = 1 mm.

Fig. 3. Typical sections of the retinas taken from newborn rat pups 1, 2, and 10 weeks after exposure to hyperoxia. These are higher magnification (10�)

images of those found in Fig. 1 and provide an image from which the retinal vascular density may be calculated. (A–C) Age-matched controls. (D) 1 ATA

group, 1 week. (E) 1 ATA, 2 weeks. (F) 1 ATA group, 10 weeks. (G) 1.5 ATA group 1 week. (H) 1.5 ATA group, 2 weeks. (I) 1.5 ATA group, 10 weeks. (J) 3.0

ATA group, 1 week. (K) 3.0 ATA group, 2 weeks. (L) 3.0 ATA group, 10 weeks. Scale bar = 50 Am.

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161154

70.497% after administration of 100% oxygen at 3.0 ATA,

1.5 ATA, and 1.0 ATA, respectively.

Retinal vascular density

We next tested if the hyperoxia that produced neuro-

protection would result in retinopathy of prematurity. Ret-

inal vasculature of newborn rat pups exposed to hyperoxia

Table 1

Structural assessment of retina after hyperoxia exposure

Groups INL

(no. of cells)

INL no./area

(no./Am2)

OPL thickne

(Am)

Control 2 weeks 118 F 7.077 0.0127 F 0.000583 16.527 F 0.

1 ATA 2 weeks 112 F 3.228 0.0136 F 0.000322 16.882 F 0.

1.5 ATA 2 weeks 123 F 1.350 0.0144 F 0.000512 16.387 F 0.

3.0 ATA 2 weeks 117 F 1.802 0.0133 F 0.000577 16.759 F 0.

ROP 2 weeks 106 F 1.951 0.0166 F 0.00147 9.150 F 0.

Control 10 weeks 116 F 4.655 0.0130 F 0.000429 16.068 0.381

1 ATA 10 weeks 121 F 1.562 0.0131 F 0.000715 16.718 0.232

1.5 ATA 10 weeks 121 F 1.280 0.0134 F 0.000121 16.195 0.345

3.0 ATA 10 weeks 122 F 1.822 0.0124 F 0.000269 16.777 F 0.

Data are represented as mean F SEM. INL indicates inner nuclear layer; OPL, out

*P < 0.001 (ANOVA) compared to all other groups in 2-week time point.

at various pressures (1, 1.5, and 3 ATA) is shown in Fig. 2.

Higher magnification (10�) sections of the retinas (Fig. 3)

were used in the calculation of retinal vascular density for

each group (Table 1). No differences were observed in the

retinal flat mounts at low magnification or high magnifica-

tion and retinal vascular density calculations showed no

difference among the control, 1.0, 1.5, and 3.0 ATA groups

at 1 (data not shown), 2, and 10 weeks after hyperoxia

ss ONL

(no. of cells)

ONL no./area

(no./Am2)

RVD

115 317 F 0.391 0.0229 F 0.000930 0.0421 F 0.00596

286 326 F 9.426 0.0258 F 0.000521 0.0463 F 0.00252

187 331 F 8.161 0.0259 F 0.000589 0.0383 F 0.00374

168 325 F 6.423 0.0238 F 0.000146 0.0416 F 0.00410

453* 334 F 6.001 0.0281 F 0.00119 0.110 F 0.00667*

336 F 9.346 0.0241 F 0.000103 0.0457 F 0.00304

335 F 6.624 0.0242 F 0.000147 0.0444 F 0.0163

336 F 4.809 0.0246 F 0.000568 0.0442 F 0.00460

410 332 F 2.517 0.0242 F 0.000328 0.0436 F 0.00524

er plexiform layer; ONL, outer nuclear layer; RVD, retinal vascular density.

Fig. 4. Retinal vasculature of newborn rat pups 2 weeks after exposure to 100% oxygen at 3 ATA for 10 h. (A) Retinal flat-mount of age-matched control. (B)

Retinal flat-mount of pup from retinopathy of prematurity group. (C) Inversion of image in A. (D) Inversion of image in B. (E) Typical section of retina from an

age-matched control pup at higher magnification (10�). (F) Typical section of retinal from a pup from retinopathy of prematurity group at higher magnification

(10�). Note the increase in vasculature seen in the retina from the retinopathy of prematurity pup (B, D, and F). Scale bar = 1 mm in A–D. Scale bar = 50 Amin E–F.

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161 155

exposure. We did expose some pups to 100% oxygen at 3

ATA for 10 h to induce retinopathy. Newborn rat pups in

this retinopathy of prematurity group did show signs of

increased vascular proliferation in their retinas 2 weeks after

hyperoxia exposure when compared to age-matched con-

trols (Fig. 4). There was also a significant increase (P <

0.001, ANOVA) in retinal vascular density values for the

retinopathy of prematurity group at 2 weeks compared to the

other groups at the same time point (Table 1).

Histological analysis

The effects of hyperoxia on the structure of the retina

were further assessed on serial sections of the retina (Fig.

5). The following parameters were measured: the number

of cells in the inner nuclear layer (INL) and outer nuclear

layer (ONL), the thickness of the outer plexiform layer

(OPL) (Am), and the area of both the INL and ONL (Am2)

(Table 1). The number of cells per unit area was also

determined. No difference was found in the number of cells

in the INL, the thickness of the OPL, and the number of

cells in the ONL among the control, 1, 1.5 and 3.0 ATA

groups at both 2 and 10 weeks. Retinas of newborn pups in

the retinopathy of prematurity group (Fig. 4) had an OPL

that had a thickness of 9.150 F 0.453, which was signif-

icantly (P < 0.001, ANOVA) thinner (arrow Fig. 4) than

the OPL of the age-matched control group as well as the 1,

1.5, and 3.0 ATA groups at 2 weeks after hyperoxia

exposure. There were no differences in any of the other

parameters measured.

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161156

Fig. 6. HIF-1a protein expression 24 h after hyperoxia exposure. (A) Age-matched control, arrow denotes positive HIF-1a staining. (B) 1 ATA group.

(C) 1.5 ATA group. (D) 3.0 ATA group. (E) retinopathy of prematurity group, arrow denotes HIF-1a expression in the inner limiting membrane. 100�,

Scale bar = 10 Am.

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161 157

HIF-1a and VEGF protein expression

HIF-1a and VEGF protein expressions were assessed

in the retinas of newborn rat pups 24 h after hyperoxia

exposure. HIF-1a protein expression (Fig. 6) was evident

in retinas of pups in all groups examined. Labeling was

observed at the level of the inner retinal surface near the

nerve fiber layer and within the INL. No differences in

HIF-1a protein expression were noted between the con-

trol, 1, 1.5, and 3.0 ATA groups. However, HIF-1a

protein expression pattern was different in the retinas of

retinopathy of prematurity pups. The majority of HIF-1a

Fig. 5. Hematoxylin and eosin staining of retinas from newborn rat pups 2 and 10 w

group, 2 weeks. (D) 1 ATA group, 10 weeks. (E) 1.5 ATA group, 2 weeks. (F) 1.5 A

(I) retinopathy of prematurity group, 2 weeks. Note the thinning of the OPL in the

panel A: INL, inner nuclear layer; ONL, outer nuclear layer; OPL, outer plexifor

section. Scale bar = 20 Am.

expression was found to be in the inner limiting mem-

brane (Fig. 6E, arrow), which is the layer closest to the

retinal vessels and there was also less expression evident

in the inner retinal surface near the nerve fiber layer and

within the INL. Labeling of VEGF (Fig. 7) was also

observed at the level of the inner retinal surface near the

nerve fiber layer (NFL) and within the INL. No differ-

ences in the extent of expression of VEGF were noted

between the control group, 1, 1.5, and 3.0 ATA groups.

VEGF protein expression in the retinopathy of prematurity

group, like the HIF-1a protein expression, was also found

to be in the inner limiting membrane with more intense

eeks after exposure to hyperoxia. (A–B) Age-matched controls. (C) 1 ATA

TA group, 10 weeks. (G) 3.0 ATA group, 2 weeks. (H) 3.0 ATA, 10 weeks.

retina from the retinopathy of prematurity group (I, arrow). Abbreviations in

m layer. Schematic in right lower corner shows regions analyzed for each

Fig. 7. VEGF protein expression 24 h after hyperoxia exposure. (A) Age-matched control group, arrow denotes positive VEGF staining. (B) 1 ATA group.

(C) 1.5 ATA group. (D) 3.0 ATA group. (E) Retinopathy of prematurity group, arrow denotes VEGF expression in the inner limiting membrane. 100�,

Scale bar = 10 Am.

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161158

staining seen (Fig. 7E, arrow) and there was less expres-

sion evident in the inner retinal surface near the NFL and

within the INL.

Discussion

The retina appears to only be susceptible to oxidant stress

during developmental stages because maturational events

occurring around the 36th postconceptual week in humans

and the third postnatal week in rodents make the vessels

resistant to hyperoxia-induced obliteration (Brooks et al.,

2001). If the use of supplemental oxygen therapy (pressur-

ized or not) is to be used in the treatment of premature or

hypoxic infants, questions concerning the safety of hyper-

oxia, especially in regards to the retina, need to be

addressed. We show here, that hyperoxia at pressures

ranging from 1.0 ATA to 3 ATA is neuroprotective when

administered after a hypoxic– ischemic insult and that

hyperoxia, as used in this study, does not cause the struc-

tural changes or abnormal vascularization in the retina that

are associated with retinopathy of prematurity. We also

found that there are not any differences in the expression

of HIF-1a or VEGF in the retinas of pups exposed to

hyperoxia when compared to control pups. Thus, a single

1-h application of oxygen, either normobaric or hyperbaric,

appears to be a safe treatment protocol for neonates after a

hypoxia–ischemia insult.

Hypoxic conditions can trigger neovascular proliferation

through the induction of potent angiogenic factors such as

VEGF (Ricci et al., 2000), which is regulated by HIF-1a

under hypoxic conditions (Miyamoto et al., 2002). During a

hypoxic– ischemic insult, the HIF-1a-VEGF system is

thought to provide protective measures by reconstructing

the vasculature through the development of collateral ves-

sels, thus acting as a rescue mechanism. However, this does

not appear to be the case in the eye (Miyamoto et al., 2002).

The pathophysiology of retinopathy of prematurity can be

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161 159

separated into two distinct phases. Phase I is termed hyper-

oxia-vasocessation and occurs when an infant or animal is

placed in a hyperoxic environment. The hyperoxia can

cause a downregulation of VEGF leading to the cessation

of normal retinal vascularization (Reynolds, 2001; Shih et

al., 2003). Phase II is termed hypoxia-vasoproliferation and

occurs when the infant or animal is moved from the hyper-

oxic environment back to room air causing a relative

hypoxia to develop in the retina. An increase in VEGF in

response to this hypoxia can lead to the proliferation of

vessels that can invade the vitreous and the retina appearing

as the pathological phenomena that cause the various

clinical and experimental stages of retinopathy of prematu-

rity (Provis et al., 1997; Stone et al., 1995). Subjecting mice

to 75% oxygen from postnatal day (P) 7-P12 leads to the

development of extraretinal neovascularization by P17

(Brooks et al., 2001) through an increase in VEGF protein

expression when the mice are returned to room air (Brooks

et al., 2001; Pierce et al., 1996). Our results showed that a

single exposure to hyperoxia did not result in an over-

expression of VEGF when the pups were returned to room

air for 24 h, nor did we see any indications of an increase in

retinal vascularization within 10 weeks after the exposure.

Torbati et al. (1995) have shown that retinopathy of prema-

turity can develop in newborn rats exposed to 100% oxygen

at 5 ATA for 5 h, whereas Ricci and Calogero (1988) have

shown that exposing newborn rats to 80% oxygen at 1.8

ATA for 10 days after birth did not cause retinopathy of

prematurity. However, Ricci and Calogero (1988) also

showed that exposing newborn rat pups to 80% oxygen at

normobaric pressure for 5 or 10 days after birth did cause

retinopathy of prematurity. The vasoconstrictive response

caused by hyperbarism can be used to explain the results of

both studies. Hyperbarism can constrict the choriocapillaries

and reduce the amount of oxygen transported from the

choroid to the inner retina during the period of hyperoxia.

The vasoconstrictive response will vary with the degree of

hyperbarism, so at high pressures the constriction can cause

a severe and prolonged reduction in choroidal and retinal

blood flow. On return to room air after exposure to low

levels of hyperbarism, the hypoxemic environment the

retina finds itself in will be reduced and the stimulus for

vasoproliferation will be decreased (Ricci and Calogero,

1988). On return to room air from exposure to higher levels

of hyperbarism, oxidative damage created by hypoxia–

ischemia can lead to the induction of retinal vasoprolifera-

tion (Torbati et al., 1995). These results and those of this

study suggest that it is the duration of the exposure that is

important when hyperoxia is given at normobaric pressure,

that the pressure is important when hyperoxia is given at

hyperbaric pressures, and that hyperoxic exposures can be

safe when administered at the appropriate pressure and

duration.

Exposure to hyperoxia during the first 14 days of

postnatal life in rat pups can prevent the normal develop-

ment of the OPL and result in long-lasting effects that can

hamper vision (Dembinska et al., 2001; Lachapelle et al.,

1999). The OPL is the retinal layer where synaptic contacts

between the photoreceptors and second order neurons occur

and its formation in rats begins around P5 and commences

on P12 (Lachapelle et al., 1999). Dembinska et al. (2001)

have shown that a progressive increase in the duration of

hyperoxia causes a gradual thinning of the OPL. Our results

show that a single exposure to hyperoxia at normobaric and

hyperbaric pressures did not result in the thinning of the

OPL (Table 1), suggesting that this short exposure does not

cause the structural anomalies in the retina that are associ-

ated with prolonged hyperoxia exposure.

Despite vast improvements that have been made in

neonatal intensive care practices, the incidence of retinop-

athy of prematurity has been stable over the last 2 decades

(Reynolds, 2001). The idea that elevated levels of oxygen

causes retinopathy of prematurity is a main reason why

physicians are hesitant about treating hypoxic or premature

infants with normobaric or hyperbaric oxygen, even though

a recent clinical trial showed that supplemental oxygen

given to infants with pre-threshold retinopathy of prematu-

rity did not result in a worsening of the disease (The STOP-

ROP Multicenter Study Group, 2000) and experimentally it

has been shown that supplemental oxygen therapy can

actually reduce the proliferative vasculopathy of retinopa-

thy of prematurity in kittens (Tailoi et al., 1995). This

stigma regarding supplemental oxygen has been around

since the early 1950s (Campbell, 1951; Crosse and Evans,

1952; Patz et al., 1952) when for lack of a better explana-

tion, it was thought that the oxygen used to treat the infants

was causing the blindness (Neubauer, 2002) even though

retinopathy of prematurity can occur in term and pre-term

infants not exposed to elevated levels of oxygen (Lucey and

Dangman, 1984). Retinopathy of prematurity shares its

pathophysiological characteristics with other common ocu-

lar diseases such as diabetic retinopathy, central vein

occlusion, and age-related macular degeneration (Mechou-

lam and Pierce, 2003), all of which share the common

feature of the development of a somewhat hypoxic state

(Frank, 2004). The use of oxygen in neonatal intensive care

units in the care of premature infants normally includes

keeping the infants on ventilators for days to weeks (Sinha

and Tin, 2003) which will certainly result in hypoxic

conditions when the infants are taken off the ventilators.

Given that long-term exposures to elevated levels of oxy-

gen can result in the onset of a hypoxic state on return to

normal room air, the incident of the hyperoxia–hypoxia

swings needs to be managed accordingly with possibly

some sort of step down strategy that will prevent the onset

of severe hypoxia. Recently, Chow et al. (2003) have

reported that through a strict management of oxygen

delivery and monitoring the incidence of retinopathy of

prematurity has decreased significantly in their neonatal

intensive care center. The authors also reported that the

strict guidelines of the program allowed for the monitoring

of oxygen levels to avoid the repeated episodes of hyper-

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161160

oxia–hypoxia in very low birthweight infants (Chow et al.,

2003). In light of the potential toxic effects and caution

surrounding the use of elevated levels of oxygen at normo-

baric pressures, treatment with oxygen at increased pressure

may be a more suitable treatment for premature and

hypoxic infants. Mild hyperbarism, as mention above, has

been shown to be safe, in regards to the retina, when

administered to rats for 10 days (Ricci and Calogero,

1988) and HBO is being used to treat newborns and

children with radiation-induced bone and soft tissue com-

plications, cyanotic congenital heart disease, and CO poi-

soning (Ashamalla et al., 1996; Chuba et al., 1997; Rudge,

1993). As mentioned above, consideration of HBO treat-

ment for hypoxia–ischemia is derived from the belief that

HBO therapy might salvage the still viable, though non-

functioning tissue surrounding the area of insult through an

increase in the amount of oxygen available to the brain

tissue (Nighoghossian and Trouillas, 1997). Other possible

effects of HBO include the reduction of pressure within the

brain caused by swelling, the restoration of blood–brain

barrier function and cell membrane function, the scaveng-

ing of free radicals, and the reduction in the stickiness of

blood products (Neubauer, 2002). All of which aid in the

reduction of brain injury. In recent years, HBO treatment

has emerged as a treatment for cerebral palsy (CP), a

common pathology associated with a hypoxic–ischemic

insult during the perinatal period. Cases studies have shown

that administration of HBO (sometimes years after the

initial injury) to children who had experienced perinatal

hypoxia– ischemia led to the dramatic improvement of

these patients in regards to behavior and alertness (Neu-

bauer, 2002).

It needs to be pointed out that the purpose of this study

was not to induce retinopathy of prematurity in neonatal

rats (hence, we did not use the established model of

oxygen-induced retinopathy), only to test if a single

treatment of hyperoxia at various pressures would result

in the retinal changes associated with retinopathy. In

conclusion, we have shown that a single exposure to

100% oxygen for 1 h at normobaric or hyperbaric pres-

sures did not cause retinopathy in newborn rat pups,

suggesting that when used correctly hyperoxia does not

lead to the development of retinopathy of prematurity. The

results obtained in this study may help to open the door for

the use of hyperoxia therapy for the treatment of hypoxic

newborn infants since both normobaric oxygen (Singhal et

al., 2002a,b) and hyperbaric oxygen (Badr et al., 2001; Yin

et al., 2002) have been shown to be effective in neonate

and adult models of stroke.

Acknowledgments

This work was partially supported by American Heart

Association Bugher Foundation for Stroke Award and by

NIH NS45694, NS43338, and HD43120 to JHZ.

References

Ashamalla, H.L., Thom, S.R., Goldwein, J.W., 1996. Hyperbaric oxygen

therapy for the treatment of radiation-induced sequelae in children. The

University of Pennsylvania experience. Cancer 77, 2407–2412.

Badr, A.E., Yin, W., Mychaskiw, G., Zhang, J.H., 2001. Effect of hyper-

baric oxygen on striatal metabolites: a microdialysis study in awake

freely moving rats after MCA occlusion. Brain Res. 916, 85–90.

Beauchamp, M.H., Marrache, A.M., Hou, X., Gobeil Jr., F., Bernier, S.G.,

Lachapelle, P., Abran, D., Quiniou, C., Brault, S., Peri, K.G., Roberts,

J., Almazan, G., Varma, D.R., Chemtob, S., 2002. Platelet-activating

factor in vasoobliteration of oxygen-induced retinopathy. Invest. Oph-

thalmol. Vis. Sci. 43, 3327–3337.

Brooks, S.E., Gu, X., Samuel, S., Marcus, D.M., Bartoli, M., Huang,

P.L., Caldwell, R.B., 2001. Reduced severity of oxygen-induced

retinopathy in eNOS-deficient mice. Invest Ophthalmol. Vis. Sci.

42, 222–228.

Calvert, J.W., Yin, W., Patel, M., Badr, A., Mychaskiw, G., Parent, A.D.,

Zhang, J.H., 2002. Hyperbaric oxygenation prevented brain injury in-

duced by hypoxia– ischemia in a neonatal rat model. Brain Res. 951,

1–8.

Calvert, J.W., Zhou, C., Nanda, A., Zhang, J.H., 2003. Effect of hyperbaric

oxygen on apoptosis in neonatal hypoxia– ischemia rat model. J. Appl.

Physiol. 95, 2072–2080.

Campbell, K., 1951. Intensive oxygen therapy as a possible cause of retro-

lental fibroplasia; a clinical approach. Med. J. Aust. 2, 48–50.

Chow, L.C., Wright, K.W., Sola, A., 2003. Can changes in clinical practice

decrease the incidence of severe retinopathy of prematurity in very low

birth weight infants? Pediatrics 111, 339–345.

Chuba, P.J., Aronin, P., Bhambhani, K., Eichenhorn, M., Zamarano, L.,

Cianci, P., Muhlbauer, M., Porter, A.T., Fontanesi, J., 1997. Hyperbaric

oxygen therapy for radiation-induced brain injury in children. Cancer

80, 2005–2012.

Crosse, V.M., Evans, P.J., 1952. Prevention of retrolental fibroplasia. AMA

Arch. Ophthalmol. 48, 83–87.

Dembinska, O., Rojas, L.M., Varma, D.R., Chemtob, S., Lachapelle, P.,

2001. Graded contribution of retinal maturation to the development of

oxygen-induced retinopathy in rats. Invest Ophthalmol. Vis. Sci. 42,

1111–1118.

Ferriero, D.M., 2001. Oxidant mechanisms in neonatal hypoxia– ischemia.

Dev. Neurosci. 23, 198–202.

Frank, R.N., 2004. Diabetic retinopathy. N. Engl. J. Med. 350, 48–58.

Han, B.H., D’Costa, A., Back, S.A., Parsadanian, M., Patel, S., Shah, A.R.,

Gidday, J.M., Srinivasan, A., Deshmukh, M., Holtzman, D.M., 2000.

BDNF blocks caspase-3 activation in neonatal hypoxia– ischemia. Neu-

robiol. Dis. 7, 38–53.

Johnston, M.V., 2001. Excitotoxicity in neonatal hypoxia. Ment. Retard.

Dev. Disabil. Res. Rev. 7, 229–234.

Lachapelle, P., Dembinska, O., Rojas, L.M., Benoit, J., Almazan, G.,

Chemtob, S., 1999. Persistent functional and structural retinal anoma-

lies in newborn rats exposed to hyperoxia. Can. J. Physiol. Pharmacol.

77, 48–55.

Lucey, J.F., Dangman, B., 1984. A reexamination of the role of oxygen in

retrolental fibroplasia. Pediatrics 73, 82–96.

Mechoulam, H., Pierce, E.A., 2003. Retinopathy of prematurity: molecular

pathology and therapeutic strategies. Am. J. Pharmacogenomics 3,

261–277.

Mink, R.B., Dutka, A.J., 1995. Hyperbaric oxygen after global cerebral

ischemia in rabbits reduces brain vascular permeability and blood flow.

Stroke 26, 2307–2312.

Miyamoto, N., Mandai, M., Takagi, H., Suzuma, I., Suzuma, K., Koyama,

S., Otani, A., Oh, H., Honda, Y., 2002. Contrasting effect of estrogen on

VEGF induction under different oxygen status and its role in murine

ROP. Invest Ophthalmol. Vis. Sci. 43, 2007–2014.

Neubauer, R.A., 2002. New hope for the neurologically damaged child,

cerebral palsy, anoxic ischemic encephalopathy, and traumatic brain

injury. In: Joiner, J.T. (Ed.), The Proceedings of the 2nd International

J.W. Calvert et al. / Experimental Neurology 189 (2004) 150–161 161

Symposium on Hyperbaric Oxygenation for Cerebral Palsy and the

Brain-injured child. Best Publishing Company, Flagstaff, pp. 3–7.

Nighoghossian, N., Trouillas, P., 1997. Hyperbaric oxygen in the treatment

of acute ischemic stroke: an unsettled issue. J. Neurol. Sci. 150, 27–31.

Palmer, E.A., 2003. Implications of the natural course of retinopathy of

prematurity. Pediatrics 111, 885–886.

Patz, A., Hoeck, L.E., De La, C.R.U.Z., 1952. Studies on the effect of high

oxygen administration in retrolental fibroplasia: I. Nursery observa-

tions. Am. J. Ophthalmol. 35, 1248–1253.

Pierce, E.A., Foley, E.D., Smith, L.E., 1996. Regulation of vascular endo-

thelial growth factor by oxygen in a model of retinopathy of prematu-

rity. Arch. Ophthalmol. 114, 1219–1228.

Provis, J.M., Leech, J., Diaz, C.M., Penfold, P.L., Stone, J., Keshet, E.,

1997. Development of the human retinal vasculature: cellular relations

and VEGF expression. Exp. Eye Res. 65, 555–568.

Reynolds, J.D., 2001. The management of retinopathy of prematurity. Pae-

diatr. Drugs 3, 263–272.

Ricci, B., Calogero, G., 1988. Oxygen-induced retinopathy in newborn

rats: effects of prolonged normobaric and hyperbaric oxygen supple-

mentation. Pediatrics 82, 193–198.

Ricci, B., Ricci, F., Maggiano, N., 2000. Oxygen-induced retinopathy in

the newborn rat: morphological and immunohistological findings in

animals treated with topical timolol maleate. Ophthalmologica 214,

136–139.

Rice III, J.E., Vannucci, R.C., Brierley, J.B., 1981. The influence of im-

maturity on hypoxic– ischemic brain damage in the rat. Ann. Neurol. 9,

131–141.

Rudge, F.W., 1993. Carbon monoxide poisoning in infants: treatment with

hyperbaric oxygen. South. Med. J. 86, 334–337.

Shih, S.C., Ju, M., Liu, N., Smith, L.E., 2003. Selective stimulation of

VEGFR-1 prevents oxygen-induced retinal vascular degeneration in

retinopathy of prematurity. J. Clin. Invest. 112, 50–57.

Singhal, A.B., Dijkhuizen, R.M., Rosen, B.R., Lo, E.H., 2002a. Normo-

baric hyperoxia reduces MRI diffusion abnormalities and infarct size in

experimental stroke. Neurology 58, 945–952.

Singhal, A.B., Wang, X., Sumii, T., Mori, T., Lo, E.H., 2002b. Effects of

normobaric hyperoxia in a rat model of focal cerebral ischemia-reper-

fusion. J. Cereb. Blood Flow Metab. 22, 861–868.

Sinha, S.K., Tin, W., 2003. The controversies surrounding oxygen therapy

in neonatal intensive care units. Curr. Opin. Pediatr. 15, 161–165.

Smith, L.E., 2002. Pathogenesis of retinopathy of prematurity. Acta Pae-

diatr., Suppl. 91, 26–28.

Stone, J., Itin, A., Alon, T., Pe’er, J., Gnessin, H., Chan-Ling, T., Keshet,

E., 1995. Development of retinal vasculature is mediated by hypoxia-

induced vascular endothelial growth factor (VEGF) expression by neu-

roglia. J. Neurosci. 15, 4738–4747.

Supplemental Therapeutic Oxygen for Prethreshold Retinopathy Of Pre-

maturity (STOP-ROP), 2000. A randomized, controlled trial: I. Primary

outcomes. Pediatrics 105, 295–310.

Stout, A.U., Stout, J.T., 2003. Retinopathy of prematurity. Pediatr. Clin.

North Am. 50, 77–87.

Tailoi, C.L., Gock, B., Stone, J., 1995. Supplemental oxygen therapy. Basis

for noninvasive treatment of retinopathy of prematurity. Invest Ophthal-

mol. Vis. Sci. 36, 1215–1230.

Torbati, D., Peyman, G.A., Wafapoor, H., Shaibani, S.B., Khoobehi, B.,

1995. Experimental retinopathy by hyperbaric oxygenation. Undersea

Hyperb. Med. 22, 31–39.

Weinberger, B., Laskin, D.L., Heck, D.E., Laskin, J.D., 2002. Oxy-

gen toxicity in premature infants. Toxicol. Appl. Pharmacol. 181,

60–67.

Yin, W., Badr, A.E., Mychaskiw, G., Zhang, J.H., 2002. Down regulation

of COX-2 is involved in hyperbaric oxygen treatment in a rat transient

focal cerebral ischemia model. Brain Res. 926, 165–171.

Zhou, C., Li, Y., Nanda, A., Zhang, J.H., 2003. HBO suppresses Nogo-A,

Ng-R, or RhoA expression in the cerebral cortex after global ischemia.

Biochem. Biophys. Res. Commun. 309, 368–376.