The midgut of the tobacco hornworm (Manduca sexta) is a highly aerobic tissue that is destroyed by programmed cell death during larval-pupal metamorphosis. The death of the epithelium begins after commitment to pupation, and the oxygen consumption of isolated midgut mitochondria decreases soon after commitment. To assess the role of the electron transport chain in this decline in mitochondrial function, the maximal activities of complexes I–IV of the respiratory chain were measured in isolated midgut mitochondria. Whereas there were no developmental changes in the activity of complex I or III, activities of complexes II and IV [cytochrome c oxidase (COX)] were higher in mitochondria from precommitment than postcommitment larvae. This finding is consistent with a higher rate of succinate oxidation in mitochondria isolated from precommitment larvae and reveals that the metamorphic decline in mitochondrial respiration is due to the targeted destruction or inactivation of specific sites within the mitochondria, rather than the indiscriminate destruction of the organelles. The COX turnover number (e−·s−1·cytochrome aa3−1) was greater for the enzyme from precommitment than postcommitment larvae, indicating a change in the enzyme structure and/or its lipid environment during the early stages of metamorphosis. The turnover number of COX in the intact mitochondria (in organello COX) was also lower in postcommitment larvae. In addition to changes in the protein or membrane phospholipids, the metamorphic decline in this rate constant may be a result of the observed loss of endogenous cytochrome c.
- Manduca sexta
- cytochrome c oxidase
- turnover number
- programmed cell death
during the metamorphosis of holometabolous insect larvae, there is a dramatic restructuring of the body plan involving the programmed death of obsolete cells. In larval Lepidoptera, the large midgut, which is responsible for the digestion of food eaten by the ravenous caterpillar, is destined to undergo programmed cell death (PCD) and be replaced by a much smaller epithelium on metamorphosis to the nonfeeding pupal stage. The endocrine system coordinates metamorphosis; in the tobacco hornworm (Manduca sexta), this process is initiated by a small rise in hemolymph ecdysteroids (“commitment peak”) around day 4 of the fifth and final larval instar (44). After commitment, the insect stops feeding, and the rate of oxygen consumption by the midgut tissue (9), as well as the isolated midgut mitochondria, declines (6, 8, 9). Because the midgut has no anaerobic capacity (7, 8, 13), this decline in aerobic metabolism is accompanied by a decline in energy-demanding processes such as active ion transport (9). Nevertheless, sufficient energy production is maintained to ensure that the larval midgut retains its structural integrity until a continuous layer of pupal epithelial cells is formed on day 7 of the larval instar (35).
The metabolic processes responsible for the changes in midgut mitochondrial respiration during the early stages of metamorphosis have been studied using top-down control and elasticity analysis (6). This approach conceptually divides oxidative phosphorylation into three blocks of reactions: 1) the “substrate oxidation system,” which includes metabolite transporters, the citric acid cycle, and the electron transport chain (ETC); 2) the “phosphorylation system,” which includes the F1Fo ATP synthase, the phosphate transporter, and the adenine nucleotide translocase; and 3) the “proton leak,” which includes passive cation cycles across the inner mitochondrial membrane. The flux through each of these subsystems over a range of protonmotive forces is then determined. In midgut mitochondria, the substrate oxidation system exerts ∼90% of the control over state 3 respiration in mitochondria isolated from the midguts of day 2, day 4, and day 5 larvae. Elasticity analysis revealed a lower flux through the substrate oxidation system at a given protonmotive force in mitochondria isolated from day 4 larvae (around the time of commitment) and day 5 larvae (after commitment). In contrast, there is little change in the flux through the proton leak and the phosphorylation system. Elasticity analysis, however, cannot reveal which processes within a given subsystem are changed during metamorphosis, but Chamberlin (6) demonstrated changes in the levels of midgut cytochromes a and c during metamorphosis. This finding indicates that alterations in the ETC complexes may be responsible, in part, for the observed decline in the activity of the substrate oxidation system, as the programmed death of the epithelium proceeds. Although a great deal of attention has been paid to the role of cytochrome c, a mobile component of the ETC, in PCD (36), little work has focused on how the activity of the ETC complexes may be changed during the programmed destruction of cells. Ricci et al. (33), however, demonstrated that permeabilization of the outer membrane of mouse liver mitochondria with truncated BID made the mitochondria vulnerable to caspase-3, an executioner caspase in the apoptotic pathway (18). Caspase-3 inhibits the oxidation of substrates that provide electrons for complexes I or II, whereas the function of complexes III and IV was unaffected. One specific target of caspases is the p75 subunit of complex I, and when cells express a noncleavable form of this subunit, the loss of mitochondrial function normally seen during apoptosis is delayed (34).
These studies on mammalian cells and mitochondria provide evidence that specific sites within the ETC are altered during type I cell death (apoptosis), but it is not known whether such specificity takes place during type II PCD (autophagic cell death). It is very likely that the midgut of the tobacco hornworm is destroyed by type II PCD, because the larval midgut is composed of large, cytoplasm-rich cells, and autophagic cell death is responsible for ridding insects of such cells during metamorphosis (26, 43). The only evidence that the ETC activity is affected during insect autophagy comes from a histochemical study on blowfly thoracic muscle during pupal-adult metamorphosis. During the late stages of muscle degeneration, there is a loss of succinate dehydrogenase (complex II of the ETC) activity (3), but it is not known whether this reflects a specific targeting of this ETC complex or an overall destruction of mitochondria.
By monitoring the oxygen consumption and respiratory chain activity in mitochondria isolated from the midguts of the tobacco hornworm at different stages of larval-pupal development, the present study demonstrates that specific sites within the ETC are altered in an insect tissue undergoing PCD. These studies are complemented with measurements of cytochrome levels to determine whether the activities of complexes III and IV correspond to the level of these components. In addition, because the amount of complex IV [cytochrome c oxidase (COX)] is equivalent to the amount of cytochrome aa3, it is possible to determine whether the turnover number of this enzyme is altered during metamorphosis.
MATERIALS AND METHODS
Eggs were obtained from a colony maintained at Ohio University. Larvae were raised at 25°C on a 16:8-h light-dark cycle and fed an artificial diet (no. 9783, BioServ, Frenchtown, NJ). Larvae were selected and staged as described by Chamberlin (6). Day 2 larvae are precommitment, whereas day 4 larvae are assumed to have experienced the commitment peak of ecdysteroids but do not manifest any external morphological or behavioral signs of commitment. Day 5 larvae overtly display the morphology (e.g., blue dorsal vessel) and behavior (e.g., wandering) of postcommitment larvae.
Mitochondrial isolation and oxygen consumption.
Mitochondria were isolated from whole midguts, as described by Chamberlin (6). Oxygen consumption was measured with a Clark-type electrode in a temperature-controlled (25°C) chamber, and state 3 respiration was initiated with the addition of ADP (final concentration 0.5 mM). Oxidation of succinate (5 mM) or palmitoyl carnitine (0.1 mM) + malate (0.5 mM) was monitored in mitochondria (final concentration 0.5 mg/ml) diluted in reaction medium (8). To assess the activity of COX in intact mitochondria, the oxygen consumption of mitochondria (final concentration 0.1 mg/ml) diluted in reaction medium containing 7 mM ascorbate and 0.7 mM N-N-N′-N′-tetramethyl-p-phenylenediamine (TMPD) was monitored. The rate of mitochondrial ascorbate + TMPD oxidation was corrected for nonmitochondrial reduction of oxygen and hereafter is referred to as “in organello COX activity.”
Complex IV (COX) was measured in freshly prepared mitochondria. The activities of complexes I, II, and III were measured in mitochondria that had been frozen in liquid N2 and stored at −70°C. The mitochondrial suspensions were thawed and diluted twofold in 50 mM potassium phosphate (pH 7.2). The diluted suspensions were then exposed to two additional freeze-thaw cycles before analysis.
Complex I (NADH:ubiquinone oxidoreductase, E.C. 184.108.40.206) was measured in 50 mM potassium phosphate (pH 7.2) by a method modified from those described by Ragan et al. (31) and Theron et al. (41). The oxidation of NADH (0.1 mM) was monitored at 340 nm in the presence of mitochondrial extract, antimycin (4 μg/ml), 2 mM EDTA, and 60 μM ubiquinone-10 (coenzyme Q2; omitted in the control). An extinction coefficient of 6.81 optical density units (OD)/mM (31) was used to calculate activities.
Complex II (succinate:ubiquinone oxidoreductase, E.C. 220.127.116.11) was measured in potassium phosphate (pH 7.2) by following the reduction of 75 μM 2,6-dichloroindophenol at 600 nm in the presence of mitochondrial extract, antimycin (4 μg/ml), 0.01% Triton X-100, 20 μM rotenone, 1 mM EDTA, 60 μM ubiquinone-10, and 20 mM succinate (omitted in control). An extinction coefficient of 21 OD/mM (31) was used to calculate activities.
Complex III (ubiquinol:ferricytochrome c oxidoreductase, E.C. 1.10.22) was measured in 50 mM potassium phosphate (pH 7.2) by a method modified from Theron et al. (41). The reduction of cytochrome c (50 μM) was monitored at 550 nm in the presence of 1 mM potassium cyanide, 1 mM EDTA, 60 μM ubiquinol-10, and mitochondrial extract. The reaction was first determined in the absence of the mitochondrial extract to ascertain the rate of the nonenzymatic reduction of cytochrome c. An extinction coefficient of 20 OD/mM (31) was used to calculate activities.
Complex IV (COX, E.C. 18.104.22.168) was determined polarographically according to a method modified from that described by Rafael (30). Mitochondria (10 mg/ml) were diluted 100-fold in mitochondrial reaction medium containing 7 mM ascorbate, 0.7 mM TMPD, 100 μM cytochrome c, and 0.09% Triton X-100. The rate of the COX-mediated oxygen consumption reaction was corrected for oxygen consumption in the absence of mitochondria and hereafter referred to as “COX activity.”
A ratio of the activities of the different complexes within the same mitochondrial preparations was calculated by multiplying the activities (nmol product or reactant·min−1·mg protein−1) by the appropriate stoichiometric relation to yield electron flux (e−·min−1·mg protein−1). The stoichiometries are as follows: 2e−/NADH for complex I, 2e−/2,6-dichloroindophenol for complex II, 1e−/cytochrome c for complex III, and 4e−/O2 for complex IV.
The level of cytochrome heme in mitochondrial preparations was determined by the method described by Chamberlin (6). Mitochondria (10 mg/ml) were diluted by 25% with 4% Triton X-100 in 100 mM potassium phosphate (pH 7.2) and kept on ice for 10 min. The samples were centrifuged at 1,000 g to pellet any particulate matter, and the supernatant was diluted with 1% Triton X-100 in 100 mM potassium phosphate (pH 7.2) to yield a final mitochondrial protein concentration of 3.4 mg/ml. The reduced (dithionite present) minus oxidized (ferricyanide present) spectrum was determined, and the extinction coefficients for appropriate wavelength pairs (37) were used to calculate the cytochrome concentrations. The concentration of cytochrome aa3 (COX) was calculated by dividing the cytochrome a concentration by 2. Turnover of in organello COX (e−·s−1·cytochrome aa3−1) was calculated by multiplying the oxygen consumption rate in the presence of ascorbate + TMPD by 4 and dividing by 60 and the cytochrome aa3 concentration measured in the same preparations.
TMPD was purchased from ICN (Costa Mesa, CA); all other chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Ubiquinol-10 was prepared by a modification of the method described by Ragan et al. (31). Briefly, this method entailed dissolving ubiquinone-10 in 1 ml of acidified (pH 2) ethanol and reducing it with a few crystals of sodium borohydride; 3 ml of diethylether-cyclohexane (2:1, vol/vol) were added, the sample was mixed, and the upper phase was collected. Then 1 ml of 2 M NaCl was added, the sample was mixed, and the upper phase was collected and evaporated to dryness under N2. The residue was dissolved in 1 ml of acidified ethanol and stored at −20°C.
Data were analyzed by a one-way ANOVA followed by a Fisher's least squared differences multiple comparisons test. Where appropriate, paired t-tests were employed. In all instances, P < 0.05 was considered to represent a significant difference.
The characteristics of the mitochondria used in this study are shown in Table 1. Mitochondria isolated from the midguts of larvae at all stages of development were intact as indicated by the high respiratory control values obtained when palmitoyl carnitine + malate was oxidized (9.0 ± 0.6, 9.8 ± 0.6, 9.0 ± 1.0 for days 2, 4, and 5, respectively). Paired t-tests revealed that, for a given developmental stage, the state 3 respiration rate was greater in the presence of ascorbate + TMPD than in the presence of succinate or palmitoyl carnitine + malate. In addition, the rate of oxygen consumption was greater in the presence of palmitoyl carnitine + malate than that in the presence of succinate. Mitochondria isolated from the midguts of day 5 larvae had the lowest rates of state 3 oxygen consumption for all of the substrates tested. Respiration in the presence of succinate was lower in mitochondria from day 4 than day 2 larvae.
The activities of the ETC complexes are shown in Fig. 1. There were no developmental differences in the activities of complexes I and III (Fig. 1, A and C). In contrast, the activity of complex II (Fig. 1B) was 39% and 48% greater in mitochondria isolated from day 2 than day 4 and day 5 larvae, respectively. The activity of complex IV (Fig. 1D) declined as metamorphosis proceeded. COX activity was 14% and 38% lower, respectively, in mitochondria from day 4 and day 5 than day 2 larvae. The activity was significantly lower in mitochondria from day 5 than day 4 larvae.
The ratios of the activities of the complexes did not remain constant throughout development (Fig. 2). Complex II-to-complex III, complex IV-to-complex I, and complex IV-to-complex III ratios were lower in day 5 than in day 2 mitochondria. The complex IV-to-complex I ratio was also lower in day 4 than in day 2 mitochondria.
The cytochrome concentrations in midgut mitochondria are shown in Fig. 3. The level of cytochrome b, which is found in complexes II and III, was the same in mitochondria from day 2 and day 5 mitochondria and higher than in day 4 mitochondria. The level of cytochrome c1, which is found in complex III, did not change with development. Cytochrome c levels were higher in mitochondria from day 2 than day 4 or day 5 larvae, and the cytochrome a level was lowest in mitochondria from day 5 larvae.
The turnover numbers of COX in intact mitochondria oxidizing different substrates are shown in Table 2. Turnover numbers for COX were significantly lower in mitochondria isolated from day 5 than day 2 or day 4 larvae. When oxidizing succinate or palmitoyl carnitine + malate, COX is operating below its maximal in organello rate; the apparent excess capacity of in organello COX is depicted in Fig. 4.
Larval-pupal metamorphosis is initiated after the commitment peak of ecdysteroids, which occurs around day 4 of the final larval instar (44). This is the beginning of the programmed destruction of the midgut, which is accompanied by a decrease in tissue oxygen consumption (9), as well as respiration of isolated mitochondria (6, 8, 9). Earlier studies on mitochondria isolated from the posterior midgut showed that succinate, but not palmitoyl carnitine + malate, oxidation declines at day 4, whereas by day 5, oxidation of both substrates is depressed (8, 9). The present study, using mitochondria isolated from the entire midgut, confirms these observations and indicates that decline of mitochondrial respiration is due to targeting of specific pathways, rather than indiscriminate destruction of the organelle. This view is supported by the observation that only two of the four respiratory chain complexes showed altered activity during this early phase of metamorphosis. Succinate dehydrogenase (complex II) activity declines at day 4, consistent with the observation that succinate-supported respiration declines at this stage. Studies on permeabilized mammalian mitochondria or submitochondrial particles have shown that succinate dehydrogenase confers the greatest control over flux through the ETC from complexes II–IV (2, 28). If a similar situation exists in Manduca mitochondria, then the loss of complex II activity, coupled with the high degree of control over respiration conferred by the substrate oxidation system (6), could account for the developmental decline in succinate oxidation observed in intact mitochondria.
In the present study, the activity of complex III did not change during metamorphosis, and this constancy was mirrored by the unchanging concentration of cytochrome c1, one of the redox units of complex III. Similar to complex III, the activity of complex I did not change during the early stages of metamorphosis; therefore, the ratio of their activities also remained constant. These ETC complexes are thought to functionally interact, so that they act as a single enzyme (2); therefore, maintaining a constant ratio during metamorphosis may ensure that the proper interaction between these two ETC complexes is preserved. Oxidation of palmitoyl carnitine + malate yields NADH, which enters the chain at complex I, and FADH2, which enters the chain at complex III via electron-transferring flavoprotein:ubiquinone oxidoreductase. Because the activities of complexes I and III did not vary during developmental changes, it is likely that changes in other processes (e.g., palmitoyl carnitine transferase, β-oxidation, citric acid cycle, and COX) are responsible for the decline in the rate of palmitoyl carnitine + malate oxidation in day 5 mitochondria. Of these processes, only the activity of complex IV was assessed in the present study. COX activity declined at day 4 and to an even greater extent by day 5. In addition, in organello COX was significantly diminished at day 5. It is thought that COX does not confer a great deal of control over flux (∼12–17%) through the ETC (2, 4, 28) or respiration of intact mammalian mitochondria (17). If this holds true for midgut mitochondria, then it seems unlikely that the 41% decline in the in organello COX activity observed in day 5 mitochondria can account for the 47% decline in the oxidation of palmitoyl carnitine + malate. Nevertheless, because the ratio of complex IV activity to complex I and complex III activity did decrease at day 5, it will be important to determine the control exerted by the individual (or interacting) complexes over the flux through the ETC in midgut mitochondria.
The activity of COX measured in the present study is substantially higher than that measured in isolated mitochondria from other animals [e.g., orchid bee flight muscle (38), Drosophila mitochondria (11), and various tissues from the mouse (25)]. This difference is probably due to the fact that in the present study the enzyme was assayed in freshly prepared mitochondria and a polarographic, rather than spectrophotometric, method was used. The polarographic method routinely yields higher activities, because the TMPD reduces the oxidized cytochrome c while it is bound to COX, and enzyme turnover takes place more rapidly when reduced cytochrome c does not dissociate from the enzyme (10).
The COX activity was higher in mitochondria isolated from day 2 than day 4 or day 5 larvae. This difference could be due to a metamorphic loss of the enzyme from the mitochondria, inactivation of the enzyme, or a change in the microenvironment around the protein. There was a significantly lower level of cytochrome a in mitochondria from day 5 than from day 2 larvae, which could account for the loss of COX activity. On the other hand, the cytochrome a content in day 2 and day 4 mitochondria was the same; therefore, loss of COX activity at day 4 cannot be attributed to the loss of the enzyme. This view is supported by the calculation of the turnover number (kcat) for maximal COX activity (determined by dividing the mean enzyme activities in Fig. 1D by the mean cytochrome aa3 concentrations calculated from values in Fig. 3). On the basis of these calculations, the value for day 2 mitochondria is higher than that for day 4 and day 5 mitochondria: 1,225, 1,034, and 951 e−·s−1·cytochrome aa3−1, respectively. Therefore, the properties of the enzyme or its environment are changed during metamorphosis. Because COX activity is influenced by surrounding lipids (29), these developmental differences in kcat may indicate a change in inner membrane composition as metamorphosis proceeds. Such changes in membrane composition have been observed in whole body mitochondria isolated from Drosophila undergoing larval-pupal metamorphosis (23).
The maximal kcat values of COX for midgut mitochondria calculated above are based on measurements at 25°C. Suarez et al. (39) measured COX activity polarographically in detergent-treated honeybee flight muscle homogenates at the physiological temperature of 37°C and estimated that kcat of the enzyme was 965 e−·s−1·cytochrome aa3−1 (40). With the assumption of a Q10 of 2, it appears that, at a common temperature, kcat is greater in midgut COX than in honeybee flight muscle COX. Although this conclusion must be made cautiously because different conditions were used to assay midgut and flight muscle COX, the lower kcat of the honeybee enzyme is consistent with the inverse relation between kcat and average body temperature described for other enzymes (20).
COX activity in detergent-treated midgut mitochondria is eight to nine times greater than in organello COX activity. This detergent-dependent stimulation of COX activity is similar to the sevenfold increase observed in rat liver mitochondria (42). The higher kcat in detergent-treated mitochondrial membranes may be due to alteration of COX's association with other proteins or changes in protein-lipid interactions (42). Although measurement of maximal COX activity in the presence of detergents provides an upper limit on the activity of this enzyme, the in organello COX activity more closely reflects the maximal activity of the enzyme in its normal environment, the intact mitochondrion. The turnover of in organello COX was lowest in mitochondria isolated from day 5 larvae. Although changes in the protein's structure or its lipid environment could account for the low kcat, the kcat could also be affected by the loss of endogenous cytochrome c. Loss of cytochrome c from cells undergoing PCD has been shown to inhibit mitochondrial oxygen consumption, and, in some cases (5, 27), it is possible to restore oxygen consumption with the addition of exogenous cytochrome c. Further evidence that the levels of endogenous cytochrome c can affect the in organello COX comes from studies on intact fibroblast mitochondria. When fibroblasts were induced to make more cytochrome c, but not cytochrome a, the in organello COX activity increased (19).
In Manduca midgut mitochondria, kcat was approximately three to five times greater for in organello COX than for mitochondria oxidizing succinate or palmitoyl carnitine + malate. This apparent excess capacity of COX observed in the present study is higher than that in mitochondria isolated from mammalian organs (15) or Artemia cysts (32). It has been suggested that a higher apparent excess capacity is indicative of a higher oxygen affinity of the respiratory chain (14). If this is the case, Manduca mitochondria have a higher oxygen affinity than mitochondria in mammalian heart or liver. Although the Po2 of midgut cells has not been measured, the Po2 in flight muscle of flying potato hawkmoths is ∼8.6 kPa (24), which is higher than that of mammalian liver cells [∼2–3 kPa (22)] or cardiac myocytes [∼0.6–0.9 kPa (12)]. Therefore, if the tobacco hornworm midgut has a Po2 similar to that of moth flight muscle, the putative higher oxygen affinity of the Manduca respiratory chain cannot be an adaptation to a lower cellular Po2. The midgut, however, may experience disruptions in oxygen delivery during molting (16). Because the midgut has no anaerobic capacity (7, 8, 13), the supposed higher oxygen affinity of the ETC may allow oxidative phosphorylation to continue during transient periods of hypoxia that may take place during molting.
The kcat for COX in the intact day 2 midgut cell can be calculated using the cyanide-sensitive oxygen consumption of the midgut [1.63 nmol O2·min−1·mg wet wt−1 (9)], a wet weight-to-protein conversion factor [120.4 mg protein/g fresh weight (7)], and the tissue cytochrome aa3 content [0.08 nmol cytochrome aa3/mg tissue protein (6)]. On the basis of this calculation, kcat for COX in the intact midgut is 11 e−·s−1·cytochrome aa3−1. Not surprisingly, this is lower than kcat of in organello COX. In contrast to the conditions used to study isolated mitochondria, mitochondria in the intact cell would be exposed to subsaturating concentrations of substrates, and respiration may be modulated by inhibitory factors. One such factor could be nitric oxide. The gene for nitric oxide synthase has been identified in lepidopteran midgut (21), and nitric oxide has been shown to inhibit the respiration of insect mitochondria (1).
In contrast to the kcat of COX in the intact midgut cell, kcat of COX in the flight muscles of flying honeybees is very high (747 e−·s−1·cytochrome aa3−1) and near the measured maximal activity of the enzyme (40). Whether this difference between the kcat of COX in the intact midgut and in flight muscle reflects a difference in the level of control over respiration exerted by this enzyme in these insect tissues remains to be determined.
The present study is the first to monitor the changes in the activity of all four complexes of the ETC in mitochondria isolated from a tissue undergoing PCD. During the early stages of the midgut PCD, specific complexes of the ETC decline in activity, and this finding indicates that changes in mitochondrial respiration during the early stages of metamorphosis are not due to indiscriminate destruction of these organelles. The fall in the activity of complexes II and IV in mitochondria isolated from larvae at or after the commitment to pupation is consistent with studies employing elasticity analysis, which demonstrated a developmental decline in flux through the substrate oxidation system (6). Additional studies, however, must be conducted to determine how much control over respiration is exerted by these complexes and whether their control strength changes during metamorphosis. It is not known how specific complexes might be targeted for destruction or inactivation during the developmentally programmed death of the midgut. It is unlikely that caspase-3, which inactivates complex I in apoptotic cells (34), is involved, because the activity of midgut complex I did not change during the early stages of metamorphosis. In addition, the midgut likely undergoes autophagic cell death; therefore, the processes responsible for modifying ETC activity in the metamorphosing midgut may be different from those in cells undergoing apoptosis.
This work was supported by National Science Foundation Grant IBN-0131523.
The author thanks Jean Thuma for technical assistance, Marcie Reiter for assistance in caring for the insects, and J. S. Ballantyne for helpful comments on the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society