Nectar-feeding bats and hummingbirds incorporate nectar carbohydrates into the pool of metabolized substrate shortly after a nectar meal
and use both carbohydrates and lipids on their
regular diurnal activity. In contrast, lipids are
the main fuel type on migratory flights (5–7).
Nectarivorous insects oxidize a wider range of
fuel types for flight [glucose, glycogen, trehalose,
proline, phosphoarginine, and lipids (3, 8)]. Hawkmoths are known to initiate flight using carbohydrates and shift to lipid oxidation shortly after (9),
whereas fed moths tend to burn a mixture of carbohydrates and fat during long bouts of flight (10).
Intense exercise, as in the extremely high aerobic performance of flight muscles in hovering
and flying animals, produces reactive oxygen
species (ROS) that can cause oxidative damage
to the contracting myocytes (11, 12). We tested
the effect of sugar feeding on oxidative damage
in Manduca sexta (Sphingidae), a hawkmoth.
Adults feed exclusively on nectar and hover when
feeding. Sugar-fed moths flew about 70% farther
in 180 min on flight mills than unfed moths
(mean ± SD, fed, 5.8 ± 2.7 km, n = 9; unfed, 3.4 ±
1.1 km, n = 9, t = –2.4, P = 0.0334), suggesting
that they should have higher levels of oxidative
damage. As expected, fed moths did have higher
oxidative damage to muscle protein (8.162 ±
3.2 nmol of protein-carbonyls/mg of protein for
fed moths and 5.512 ± 2.7 for unfed moths; F1,27 =
5.5576; P = 0.0259; Fig. 1A). Increased protein
damage with flight is expected, as damage to
the flight muscle accumulates until the damaged
proteins are replaced or recycled. Although damage to muscle protein might affect the function
of the muscle, it is nonlethal, and in a short-lived
insect, like adult M. sexta, the accumulation of
protein damage may outweigh the high cost of
repair. By contrast, when ROS attack cell membrane fatty acids, a chain reaction of lipid peroxidation ensues in which one lipid peroxidation
event can initiate hundreds of radical reactions
(13). This leads to a severe reduction in membrane
functions and damage to the cellular lipid bilayers, which can result in cell death. In addition
to damage to membrane fatty acids, lipid peroxidation chain reactions also lead to the production of lipid radicals that, in turn, can attack
other lipids, proteins, and nucleic acids (14). In
contrast to muscle protein, sugar-fed moths had
lower oxidative damage to their muscle cell membranes (0.035 ± 0.01 mg of malondialdehyde/mg
of protein for fed moths and 0.081 ± 0.08 for
unfed moths; F1,25 = 13.154; P = 0.0013; Fig. 1B),
indicating that antioxidant activity reduces the
peroxidation of membranes in the fed moths.
Glutathione in its reduced form (GSH) is a key
antioxidant in protecting cell membranes by re-
ducing oxidative damage. GSH donates electrons
to lipid peroxide, which stops the peroxidation
chain reaction (15, 16). We found significantly
higher ratios of GSH/GSSG in fed moths (1.9765 ±
0.5 mM glutathione disulfide/mg of protein for
fed moths and 1.557 ± 0.5 for unfed moths; F1,29 =
5.146; P = 0.0309; Fig. 1C), indicating that fed
moths are better able to recycle oxidized glutathione
(GSSG) to its reduced active form (GSH) to control
oxidative damage to flight muscle membranes.
GSSG is recycled to GSH by glutathione re-
ductase using reduced nicotinamide adenine di-
nucleotide phosphate (NADPH) as an electron
donor. The main source of NADPH is the pentose
phosphate pathway (PPP). The PPP is an essen-
tial and conserved metabolic pathway in bacteria,
plants, and animals and is involved in the synthe-
sis of nucleic acid precursors (ribose-5-phosphate)
and reducing power in the form of NADPH, two
processes vital for the maintenance of cell func-
tion (17). The PPP is divided into oxidative (ox-
PPP) and nonoxidative (nonox-PPP) branches.
Ox-PPP is an irreversible pathway that produces
ribulose-5-phosphate, NADPH, and CO2 (from car-
bon atom 1 of glucose). Ribulose-5-phosphate is
a major source of metabolic precursors for bio-
synthetic processes such as nucleic acids, whereas
NADPH functions as reducing power for anabolic
process such as fatty acid synthesis, cholesterol
synthesis, and the maintenance of a pool of re-
duced glutathione. Both NADPH and glutathione
are important antioxidants used to reduce cellular
oxidative damage. The nonox-PPP branch is a
reversible pathway that interconverts pentose
phosphate and other sugar phosphates. It con-
tributes to the synthesis of ribose-5-phosphate
and redirects excess pentose phosphate toward
glycolysis (17) and the production of the insect
blood sugar trehalose (18), which is important
in enantiostasis (8).
When carbohydrates are the sole source of
metabolic fuel, for every molecule of O2 con-
sumed, one molecule of CO2 is produced. This
yields a respiratory quotient (RQ = VCO2/VO2)
of 1.0. The published RQs of many nectarivores
are greater than 1.0. Over the past 100 years,
high RQ values have been recorded in a variety of
animals, including nectar-feeding bats [RQ values
up to 1.7 (19)], sunbirds, and their convergent
New World hummingbirds [1.3 and 1.4, respec-
tively (20)]. We found that the peak RQ of
sugar-fed M. sexta ranged between 1.11 and
1.74 (n = 10, mean ± SD = 1.38 ± 0.19). Such high
values have been previously interpreted as “lipid
synthesis,” but a mechanism by which lipid syn-
thesis causes a high RQ has never been ade-
quately explained (21–23). Declining body mass
is characteristic of many adult Lepidoptera, even
when fed as adults (24), which suggests that
lipid synthesis is not significant in these insects.
To the best of our knowledge, under aerobic
conditions, these high RQ values can only be
explained by the use of the PPP and the release
of CO2 from glucose carbon atom 1 (C1) in this
metabolic pathway. We suggest that the reduc-
tion potential of the PPP can also be used to build
endogenous antioxidant potential.
We coupled real-time respirometry with real-
time d13C analysis to determine fuel oxidation
strategies (25) of fed and unfed moths at rest
and during activity. Moths (n = 12) were fed nec-
tar containing either 13C1- or 13C2-labeled glucose
(1.0 mg/ml), then placed in a metabolic chamber
at rest. It is more energetically efficient to oxi-
dize ingested sugar directly than to first convert
it to lipids, as has been shown for nectarivorous
bats and hummingbirds (5). d13C in the breath
of the fed moth started to rise immediately after
feeding, reaching a steady-state RQ within 30 to
60 min. When moths were disturbed from rest
by shaking the chamber, they activated their
flight muscles in a preflight warm-up (4), causing
an immediate drop in RQ (Fig. 2). These changes
in RQ might also reflect different metabolisms in
different organs. For example, when moths are
at rest and fed, the fat body and digestive sys-
tem are active. By contrast, when moths are flying,
flight muscles are the primary active organ. Dur-
ing postfeeding rest, more C1 is exhaled as CO2
than for the other five glucose carbons (C2-6).
This would occur if C1 were selectively released as
CO2 in the PPP (decarboxylation of 6P-gluconate,
fig. S1). This additional CO2 results in an RQ >> 1.
When the muscles were activated in the preflight
warm-up (black arrows in Fig. 2), the d13C1 de-
creased (Fig. 2A), suggesting that the 13C1 in the
734 17 FEBRUAR Y 2017 • VOL 355 ISSUE 6326
sciencemag.org SCIENCE
Fig. 1. Oxidative damage and GSH/GSSG ratio
in flight muscles in sugar-fed and starved moths.
(A) Levels of protein oxidative damage. (B) Levels of
lipid oxidative damage. (C) Ratios of reduced to
oxidized glutathione (GSH/GSSG) (mean ± SD).
1Department of Entomology, University of Arizona, Tucson, AZ,
USA. 2Department of Biology, New Mexico State University, Las
Cruces, NM, USA. 3School of Plant Sciences and the BIO5
Institute, University of Arizona, Tucson, AZ, USA.
*Corresponding author. Email: levineran1@gmail.com
