Main

Reactive oxygen species are regarded as a possible cause of many diseases. In the area of neonatal medicine in particular, reports suggest that reactive oxygen species work as a factor in the onset of such conditions as chronic lung disease (1, 2), retinopathy of prematurity (2), and necrotizing enterocolitis (3). However, there have been no in vivo or in situ reports that establish the direct participation of reactive oxygen species in the pathology of any disease.

The luciferin derivative MCLA reacts specifically with superoxide radical and singlet oxygen, and Nishida et al. (4) and Nishida (5) have demonstrated that it has a chemiluminescence through dioxetane. The ultimate goal of the present study was to elucidate the pathophysiologic significance of reactive oxygen species in newborn infants. As a basis of investigation, we examined whether reactive oxygen species, and specifically superoxide radical, are mass-produced by the lungs via resuscitation by 100% oxygen used in routine treatment.

METHODS

Newborn piglets were used as subjects within 24 h of birth. All procedures were approved by the Animal Ethics Committee at the Kagawa Medical University. After anesthesia by intramuscular administration of pentobarbital, two catheters were inserted into the umbilical artery, one deeply for continuous measurements of blood pressure and heart rate, and the other with shallow penetration as a route for extracting blood. A catheter was also inserted into the umbilical vein for infusion of 4 mL·kg−1·h−1 of maintenance fluid (Na 50 mEq/L, K 20 mEq/L, Cl 50 mEq/L, lactate 20 mEq/L, glucose 2.7%). After intubation, pancuronium bromide was administered, spontaneous respiration was stopped, and respiration was thereafter controlled by an artificial respirator. With the animals lying on their left sides, the right chest was opened, and the lung was exposed.

The sum of microluminescence on the surface of the lung was calculated each minute in continuous measurements using ARGUS 50 Image processing system from Hamamatsu Photonics (Hamamatsu, Japan). The fraction of inspired oxygen was set at 21% except during resuscitation. The respiration rate and pressure were adjusted so that the arterial oxygen tension was >60 mm Hg and the arterial carbon dioxide tension was 35–40 mm Hg.

Chemiluminescence by MCLA.

To investigate the possibility of measuring chemiluminescence in vivo and in situ using MCLA (Tokyo Kasei Kogyo Co., Ltd., Tokyo, Japan), 2 mL of MCLA (0.2 mg/mL) was given in a bolus injection in the brachial vein, after which it was administered by i.v. drip at 1 mL/h according to Takahashi et al. (6) for MCLA administration. By contrast, one animal was administered the same volume of physiologic saline solution instead of MCLA. Acute asphyxiation was accomplished with a respiratory setting of 4 min of continuous positive airway pressure at 21% oxygen concentration to induce apnea. After 4 min, resuscitation was effected by artificial respiration using 100% oxygen. Measurements (photon counts per minute) were performed continuously from 2 min before asphyxiation until 10 min after resuscitation.

Ability to produce reactive oxygen species as a function of oxygenconcentration during resuscitation.

With reference to Takahashi et al. (6), 2 mL of MCLA (0.2 mg/mL) was given by bolus injection in the brachial vein, followed by an i.v. drip at 1 mL/h. Measurements were taken for 2 min directly before acute asphyxiation, caused by stopping respiration for 4 min using continuous positive airway pressure with 21%oxygen, resulting in bradycardia. The animals were then resuscitated with an artificial respirator, and measurements were continued for 10 min afterward.

The resuscitated animals were divided into two groups, one in which the animals had been resuscitated with 100% oxygen (100% oxygen resuscitated group, n = 8), and one in which resuscitation was performed with 21% oxygen (21% oxygen resuscitated group, n = 6), to determine whether there was a difference in the amount of reactive oxygen species produced. Furthermore, the effect on the amount of reactive oxygen species produced in the lung by asphyxiation itself, whether or not there was a difference in the amount of reactive oxygen species produced, was investigated between the 100% oxygen resuscitated group and 100% oxygen administered group (n = 7), a group that was administered 21% oxygen for 6 min, then 100% oxygen without asphyxiation.

Statistics.

Two sample populations (unpaired) were evaluated with a Welch's t test. Differences were considered significant at p < 0.05.

RESULTS

Chemiluminescence by MCLA.

Measurements (photons per minute) were collected throughout all procedures for the animal given physiologic saline solution. However, chemiluminescence values of only 0–8 photons/min (mean, 2 photons/min) were observed. In contrast, LSC of approximately 40 photons/min was observed before asphyxiation in the animal given MCLA. There was a temporary drop in LSC caused by asphyxiation, with a rapid increase reaching a maximum of 580 photons/min after resuscitation with 100% oxygen had begun (Fig. 1). The results of this preliminary experiment confirmed that it was possible to measure reactive oxygen species in vivo and in situ in newborn piglets using this method.

Figure 1
figure 1

Chemiluminescence from the lung surface. In the saline-administered animal, the LSC of only 0–8 photons/min (mean, 2 photons/min) was observed, but in the MCLA-administered animal, the LSC of approximately 40 photons/min was observed before asphyxiation, with a rapid increase reaching a maximum of 580 photons/min after resuscitation with 100% oxygen had begun.

Full size image

Ability to produce reactive oxygen species as a function of oxygenconcentration during resuscitation.

Before asphyxiation, LSC was observed in both animal groups, although the LSC was very low. LSC decreased with the elapse of time after inducing asphyxiation. In the 100% oxygen resuscitated group, LSC rose steadily after resuscitation, exhibiting maximum luminescence of 2–6 min after resuscitation had begun, when the highest amount of reactive oxygen species was produced. In the 21% oxygen resuscitated group, maximum luminescence was seen between 2 and 6 min after resuscitation had begun, but the luminescence was less than that of the 100% oxygen resuscitated group (Fig. 2). A comparison of maximum luminescence was made between the two groups by comparing the maximum chemiluminescence relative to the LSC before asphyxiation (maximum LSC after resuscitation minus LSC before asphyxiation) of both groups. The mean relative maximum chemiluminescence was 59.8 ± 39.0 photons/min (range, 11.0 to 109.5 photons/min) for the 21% oxygen resuscitated group, and 186.1 ± 72.5 photons/min (range, 105.0 to 297.5 photons/min) for the 100% oxygen resuscitated group. There was a significant difference (p = 0.0015) between the two groups (Fig. 3).

Figure 2
figure 2

Relative photon count changes and time. Average relative photon counts at each recording time are shown for both 21% O2 resuscitated and 100% O2 resuscitated groups. Data are shown as mean ± SD.

Full size image
Figure 3
figure 3

Comparison of relative maximum chemiluminescence. The mean relative maximum chemiluminescence (photons/min) is calculated as the maximum LSC after resuscitation minus the LSC before asphyxiation.

Full size image

In the 100% oxygen administered group, there was a gradual rise in LSC because of the administration of oxygen, which reached a maximum value between 2 and 5 min. The mean relative maximum luminescence for this group was 108.4 ± 39.8 photons/min (range, 59.5 to 175.5 photons/min) and was significantly higher than the 21% oxygen resuscitated group (p = 0.047), but significantly lower, at about 60%, compared with that of the 100% oxygen resuscitated group (p = 0.024), demonstrating that the production of reactive oxygen species increased significantly because of asphyxiation (Figs. 3 and 4).

Figure 4
figure 4

Photon counts from individual piglet. First, we observed photon counts from a piglet breathing 21% oxygen and 100% oxygen. The same piglet was ventilated using 21% oxygen for a sufficient time, and then we observed photon counts from the piglet after it experienced an asphyxiation load and resuscitation using 100% oxygen.

Full size image

Among the three groups, blood pressure and heart rate were compared before asphyxiation load, at the time of resuscitation (in the 100% oxygen administered group, the time when the change was made to 100% oxygen), and at the time of relative maximum chemiluminescence after resuscitation. A difference of heart rate was observed between the 100% oxygen resuscitated group and the 100% oxygen administered group before the start of asphyxiation, although the cause of this difference is unclear; no significant difference was evident after asphyxiation (Table 1

Table 1 Physiologic variables *p < 0.05. Values are mean ± SD. In the 100% oxygen group, the time of resuscitation is the time at which we changed inspiratory oxgen from 21% oxygen to 100% oxygen. Abbreviation: resus., resuscitation.
Full size table

DISCUSSION

With the advances in neonatal treatment of recent years, the survival rate of extremely low birth weight infants of <1000 g has been increasing rapidly, and now survival has also become possible for newborns with a gestational age of <24 wk. One consequence of this is that the prevention of such diseases as chronic lung disease, consisting mainly of bronchopulmonary dysplasia, and retinopathy of prematurity has become increasingly important.

It is thought that reactive oxygen species play a role in causing these diseases (1, 2). The life-span of all reactive oxygen species is extremely short, and therefore determining it in vivo is extremely difficult. Some reports have demonstrated the direct involvement of reactive oxygen species in the pathology of diseases (6–10). In the present study, MCLA, which reacts specifically with superoxide radical and singlet oxygen, was used under the same conditions in which we attempt to resuscitate in routine daily practice, to investigate the amount of reactive oxygen species produced in vivo and in situ. MCLA is a luciferin derivative of which excited species exhibit a maximum chemiluminescence of 460 nm when returning to their ground state after passing through dioxetane and in the presence of superoxide radical and singlet oxygen. A characteristic of the ARGUS 50 is that, in the range of wavelengths (280–650 nm) at which chemiluminescence can be observed, its maximum sensitivity is in the vicinity of 400 nm, and individual photons can be captured and photon-counting images calculated two-dimensionally, enabling measurement of microluminescence. With reference to the methods of Takahashi et al. (6), the present study used a model using newborn piglets within 24 h of birth, under the same conditions of resuscitation with 100% oxygen that are used regularly, in such cases as asphyxiated newborns and apnea attack in premature infants, to demonstrate in vivo and in situ production sites of large amounts of reactive oxygen species. The LSC observed before acute asphyxiation may be considered to be mainly chemiluminescence dependent on MCLA in the peripheral blood, according to the study of Takahashi et al. (6). In the current study, we found that significantly more reactive oxygen species were produced in the 100%oxygen resuscitated group than in the 100% oxygen administered group without asphyxiation load, and these results are believed to be significant in that they corroborate that large amounts of reactive oxygen species are produced in situ not only by 100% oxygen, but also by the asphyxiation load.

The main mechanisms of reactive oxygen species production during ischemic-reperfusion include 1) the production of superoxide radical by xanthine oxidase, 2) the metabolic dissociation of protoplasm and mitochondria through the incomplete reduction of oxygen, 3) the liberation of iron accompanying acidosis, and 4) the production of superoxide radical from peripheral neutrophilic leukocytes (11). Some possibilities for the disturbance in reactive oxygen species are that the production of reactive oxygen species surpasses the tissue's ability to eliminate it or that there is a failure in that ability itself, or a combination of both. Nakamura and Nishida (12, 13) and Nishida et al. (14) have investigated the rate of neutrophilic leukocyte superoxide radical production in asphyxiated neonates using a luciferin derivative-dependent chemiluminescence, and they reported differences between normal newborns and healthy adults, who have an early extra peak (first peak) that makes it bimodal. Using a tumor necrotizing factor, this first peak is the same as that seen during priming, and it has been reported that neutrophilic leukocytes are primed through cytokines from asphyxiation. It has also been reported that the ability of neutrophilic leukocytes in newborn infants to produce superoxide radical is accentuated compared with that in adults (15). The ability to eliminate reactive oxygen species has also been studied developmentally. Tanswell and Freeman (16) have reported that in rats still in the womb, the activation of SOD, catalase, and glutathione peroxidase rises with development. It has been reported by Keeney et al. (17) that in neonatal rats, the administration of oxygen for 72 h increases antioxidant enzyme activities, in particular SOD, to rise more than in adult rats, and Frank and Sosenko (18) have demonstrated that antioxidant enzyme activities in the preterm rabbit lung are lower than in the term lung and fail to increase in response to hyperoxic exposure. One may therefore assume that the ability to eliminate reactive oxygen species is poorer in premature infants than in adults and term infants. In an investigation using postmortem examinations of human lungs with a fetal age of 10–40 wk, the authors found no clear difference in SOD activity in a comparison between fetuses that had never breathed and newborns that had survived and breathed for a certain period, showing that no induction had occurred. With regard to glutathione peroxidase activity in the lungs, however, previous reports suggest that induction caused by oxygen occurs from wk 20 through 25 (19).

The results of the present experiment leave unresolved the questions of where and how the production of reactive oxygen species is brought about, and therefore these questions require further study. It has been demonstrated, however, that current methods used for resuscitation produce large amounts of reactive oxygen species, and therefore ventilation and especially resuscitation by 100% oxygen may represent a potential danger.