Biological Reviews
Volume 97, Issue 6 pp. 2106-2126
Original Article
Full Access

Avian adjustments to cold and non-shivering thermogenesis: whats, wheres and hows

Punyadhara Pani, 

Punyadhara Pani

School of Biotechnology, KIIT University, Bhubaneswar, Odisha, 751024 India

Search for more papers by this author
Naresh C. Bal, 

Corresponding Author

Naresh C. Bal

School of Biotechnology, KIIT University, Bhubaneswar, Odisha, 751024 India

Author for correspondence (Tel.: +91 674 2725466; E-mail: [email protected]).

Search for more papers by this author
First published: 28 July 2022
https://doi.org/10.1111/brv.12885
Citations: 9

ABSTRACT

Avian cold adaptation is hallmarked by innovative strategies of both heat conservation and thermogenesis. While minimizing heat loss can reduce the thermogenic demands of body temperature maintenance, it cannot eliminate the requirement for thermogenesis. Shivering and non-shivering thermogenesis (NST) are the two synergistic mechanisms contributing to endothermy. Birds are of particular interest in studies of NST as they lack brown adipose tissue (BAT), the major organ of NST in mammals. Critical analysis of the existing literature on avian strategies of cold adaptation suggests that skeletal muscle is the principal site of NST. Despite recent progress, isolating the mechanisms involved in avian muscle NST has been difficult as shivering and NST co-exist with its primary locomotory function. Herein, we re-evaluate various proposed molecular bases of avian skeletal muscle NST. Experimental evidence suggests that sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) and ryanodine receptor 1 (RyR1) are key in avian muscle NST, through their mediation of futile Ca2+ cycling and thermogenesis. More recent studies have shown that SERCA regulation by sarcolipin (SLN) facilitates muscle NST in mammals; however, its role in birds is unclear. Ca2+ signalling in the muscle seems to be common to contraction, shivering and NST, but elucidating its roles will require more precise measurement of local Ca2+ levels inside avian myofibres. The endocrine control of avian muscle NST is still poorly defined. A better understanding of the mechanistic details of avian muscle NST will provide insights into the roles of these processes in regulatory thermogenesis, which could further inform our understanding of the evolution of endothermy among vertebrates.

I. INTRODUCTION

Survival in the cold is a complex phenomenon and myriad adaptive mechanisms are observed in different animals that contribute to their evolutionary success. The maintenance of a relatively stable body temperature (Tb) in cold environments is a remarkable attribute of some vertebrates and likely was a catalyst in the Cenozoic radiation of birds and mammals. Upon acclimation to cold or acclimatization to winter conditions, birds exhibit several phenotypic and physio-metabolic adjustments that help them to minimize heat loss, reducing their thermogenic demands and thereby requirement for food (energy). Endothermic homeotherms including birds possess a unique ability to modulate heat production inside their body, which encompasses shivering as well as non-shivering thermogenesis (NST), enabling them to maintain a relatively constant Tb.

Research on NST has intensified during the last two decades with attempts to identify targetable mechanisms to counter metabolic syndromes. Classically, brown adipose tissue (BAT) has been considered the major site of NST, although skeletal muscle has long been proposed as an alternative. Landmark discoveries in recent years that ignited NST research include: characterization of activatable BAT in adult humans (Virtanen et al., 2009; McNeill, Suchacki & Stimson, 2021; Pan et al., 2020); identification of cells in human white adipose tissue (WAT) depots with BAT-like properties termed ‘beige’ fat (Wu et al., 2012; Rabiee, 2020; Singh et al., 2020); the description of uncoupling protein 1 (UCP1)-independent mechanisms of thermogenesis in BAT (Ikeda & Yamada, 2020; Roesler & Kazak, 2020; Chouchani, Kazak & Spiegelman, 2019; Verkerke & Kajimura, 2020); and delineation of futile Ca2+-cycling mechanisms in skeletal muscle as a potential contributor to NST (Rowland, Bal & Periasamy, 2015; Nowack et al., 2019; Bombardier et al., 2013; Kaspari et al., 2020; Rotter et al., 2018; Nie et al., 2017; Li et al., 2021). Although BAT is a sophisticated thermogenic mechanism found in some eutherian mammals (Nechad, 1986; Saito, Saito & Shingai, 2008), there are several mammalian clades, including marsupials and monotremes, that lack BAT yet are endothermic (Polymeropoulos, Jastroch & Frappell, 2012; Hayward & Lisson, 1992; Rose et al., 1999; Grigg, Beard & Augee, 2004). This calls for careful reconsideration of skeletal muscle as an important site of NST, for which birds can serve as an appropriate model. Birds lost the UCP1 gene from their genome early during their evolution, probably in the Jurassic period, and completely lack BAT-like structures (Johnston, 1971; Saarela et al., 1991; Walter & Seebacher, 2009; Nowack et al., 2017; Bicudo, Vianna & Chaui-Berlinck, 2001; Newman, Mezentseva & Badyaev, 2013). However, most avian clades have mechanisms to maintain Tb and many exhibit remarkable cold adaptability. For example, black-capped chickadees (Poecile atricapillus), American goldfinches (Spinus tristis) and snow buntings (Plectrophenax nivalis) can remain euthermic even at ambient temperatures much lower than those experienced under natural conditions (Cooper & Swanson, 1994; Liknes, Scott & Swanson, 2002; Vézina et al., 2020; Mayer, Lustick & Battersby, 1982; Andreasson, Nord & Nilsson, 2020). Moreover, several avian groups have radiated successfully into cold climates (Aulie, 1976; Stokkan, Mortensen & Blix, 1986; Wilson et al., 1989; Blix, 2016). Therefore, birds provide an excellent opportunity to explore UCP1/BAT-independent NST mechanisms.

The endurance capacity of endotherms to cold depends on their ability to sustain elevated levels of thermogenesis to counteract heat loss depending on their surface area to volume ratio. The avian body is adapted for flight, including both anatomical and biochemical specializations. Several of these features (such as a high muscle to body mass ratio and excellent insulation) assist in maximizing thermogenesis while minimizing heat loss (Newman et al., 2013). Using thermogram studies in birds during cold exposure, it was found that thinly feathered or bare areas, like the ophthalmic region, which has a critical role in cooling in hot climates, are the major sites of heat loss (Veghte & Herreid, 1965; Hill, Beaver & Veghte, 1980; Frost, Siegfried & Greenwood, 1975; Midtgård, 1983, 1984; Johansen & Bech, 1983; Marsh, Dawson & Wang, 1989; Chou & Guy, 1985). Also, most avian species maintain a higher basal metabolic rate (BMR, see Table 1 for glossary) than mammals of equivalent size and can sustain an elevated metabolic rate during cold exposure (Block, 1994; Klaassen, Oltrogge & Trost, 2004; Raichlen et al., 2010). For example, winter-acclimatized American goldfinches can remain normothermic and maintain a metabolic rate five times that of BMR for 8 h at −70 °C (Dawson & Carey, 1976). But the molecular mechanisms involved in cold-induced elevated metabolic rate versus NST have been difficult to differentiate in birds. Due to this ambiguity, some researchers have even suggested that avian adaptive thermogenesis is completely reliant on shivering and may not have a NST component (Aulie, 1976; West, 1965; Hohtola, 1981, 1982; Aulie & Tøien, 1988). However, early studies showed that cold (1 °C) exposure increased heat production in spinal cord-denervated pigeons (Columba livia) (which cannot shiver) indicating the presence of NST (Kayser, 1928). Several subsequent studies also indicated the existence of NST in birds and attempted to delineate the mechanisms involved (Newman et al., 2013; Duchamp & Barre, 1993; Carey et al., 2019; Rowland et al., 2015; Nowack et al., 2017; Grigg et al., 2022). Further, birds overwintering at northern latitudes show substantial increases in thermogenic capacity. Shivering can be diminished after prolonged cold acclimation, indicating greater heat production despite less muscular activity presumably due to either a greater contribution from NST or more efficient shivering thermogenesis (Bicudo et al., 2001, 2010; Nowack et al., 2017; Bal et al., 2016; Andreasson et al., 2020).

Table 1. Glossary of terms
BMR Basal metabolic rate includes all the energy (ATP/GTP) utilized by the body at a thermoneutral temperature. This definition of BMR is most suitable for homeotherms that maintain their body temperature in a precise range. In the case of heterothermic animals, BMR can be reduced by turning off thermogenic processes. For these animals, BMR might vary depending on their body temperature regimen whether homeothermic or heterothermic.
Obligatory thermogenesis Heat production inside the body that cannot be regulated. For example, when splitting energy-rich molecules such as ATP, part of the energy will inevitably be released without being coupled to the reaction being catalysed.
Regulatory thermogenesis Heat production inside the body that can be turned on or off according to thermoregulatory requirements. It can be considered an effective way to regulate energy balance by increasing/decreasing energy expenditure rapidly in response to cold.
Non-shivering thermogenesis (NST) Heat production inside the body by mechanisms that are not coupled to muscle contraction. The most established muscle-mediated mechanism of NST in mammals is based on Ca2+ slippage by sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) and is regulated by the small single transmembrane protein sarcolipin (SLN).
Isometric shivering Heat production in the muscle without visible contraction of the muscle. It is proposed to be similar to isometric muscle tremor. However, during isometric shivering myosin ATPase is fully engaged and ATP utilization by it must be minimal. Among the mechanisms contributing to heat production, SERCA-based ATP utilization might be the most important as cytosolic Ca2+ is maintained at high levels to keep the myosin–actin cross-bridges locked. Thus, isometric shivering is theoretically similar to muscle NST at least in terms of SR-based Ca2+-transport and ATP utilization.

The purpose of this review is to draw attention to the molecular mechanisms of avian thermogenesis, especially NST. We discuss recent advances in our understanding of mechanisms associated with the sites of avian NST. We critically analyse existing data and ideas proposed for muscle NST in birds in light of recent discoveries in mammals to identify gaps in knowledge. We also evaluate the heat conservation and behavioural strategies of birds to maximize their survival in cold climates. We do not discuss some aspects that may influence cold adaptation in birds, such as migration, domestication and parental care.

II. COLD ADAPTATION

Birds have evolved several strategies to overcome cold weather; these are especially prominent in birds inhabiting colder climates. With the efficient thermal insulation provided by feathers, even tiny birds like black-capped chickadees (weighing less than 30 g) can maintain a Tb of 41 °C even at ambient temperatures as low as −20 °C (Lewden et al., 2014). Birds also use behavioural and morphological/anatomical adaptations that enable heat conservation to reduce their thermogenic demand.

(1) Behavioural strategies

Birds exhibit a variety of behavioural strategies that enable heat conservation.

(a) Huddling and crowding

Some birds crowd and/or flock together to share body heat and minimize heat loss, for example tree swallows (Tachycineta bicolor), crows (Corvus spp.), ravens (Corvus corax) and ring-billed gulls (Larus delawarensis) (French, 1998; Heinrich, 2014; Arnold & Oswald, 2018; Whittow, 1986). Many birds like nuthatches (Sitta spp.), tufted titmice (Baeolophus bicolor) and downy woodpeckers (Picoides pubescens) roost in tree cavities or boxes during cold weather, while sage grouse (Centrocercus urophasianus) roost in cavities in the snow (Caorsi et al., 2019; Cooper, 1999; Dhondt, Kempenaers & De Laet, 1991; Back, Barrington & McAdoo, 1987). Interestingly, some woodpeckers excavate cavities for roosting and orient the entrance away from the prevailing wind direction in the winter but not in summer (Peterson & Grubb, 1983; Jackson & Jackson, 2004; Mainwaring, 2011).

(b) Posture

In response to cold temperatures, birds can use postural changes to decrease convective heat loss. Birds often tuck their heads under their wings (Yorzinski et al., 2018; Ryeland, Weston & Symonds, 2019), or retract their legs into their body feathers in cold weather, during flight or when sitting (Yorzinski et al., 2018; Carr & Lima, 2012). Such thermoregulatory postures conserve body heat by using warm air trapped in feathers to reduce thermogenic demand (Dawson & Whittow, 2000; Wolf & Walsberg, 2000; Carr & Lima, 2012). Distinct postures can provide thermal benefits due to reductions in exposed surface area, thereby minimizing heat loss especially from the legs and feet (Martineau & Larochelle, 1988; Ward et al., 1999). The depth of the feather layer can be increased by ptiloerection (fluffing up) of feathers, a strategy used especially during sleep (Hohtola, Rintamäki & Hissa, 1980). Such postural changes allow birds (particularly those of small size) to remain active while exposed to low temperatures or high wind chill (Walsberg, 1986; Bakken, Murphy & Erskine, 1991; Zerba, Dana & Lucia, 1999). The most thermally advantageous postures in dark-eyed juncos (Junco hyemalis) were typically used only when air temperature fell below −10 °C (Carr & Lima, 2012). It may be possible that heat-conserving postures facilitate an increase in metabolic rate during cold adaptation, but underlying mechanisms are yet to be explored. These postural adaptations may represent a constraint on the ability to initiate flight, which could increase predation risk. Hence, the use of heat-conserving postures in cold conditions may result in a trade-off between thermogenic requirements and predation risk (Cresswell, 1993).

(c) Facultative hypothermia

Facultative hypothermia involves a reduction in the metabolic rate to conserve energy. Several bird species including chickadees (Paridae), doves (Columbidae), hummingbirds (Trochilidae), mouse birds (Coliidae), tawny frogmouth (Podargus strigoides), nighthawks (Chordeilinae), blue tit (Cyanistes caeruleus), poorwill (Phalaenoptilus nuttallii) and road runners (Geococcyx spp.) show an increased reliance on facultative hypothermia during cold periods (Körtner, Brigham & Geiser, 2000, 2001; Doucette et al., 2011; Ruf & Geiser, 2015; Geiser, 2020). Hypothermia and torpor are likely to be beneficial when the food supply is restricted, but may have associated costs such as impaired quality of sleep (Mueller, Steinmeyer & Kempenaers, 2012), impaired immune function (Nord, Hegemann & Folkow, 2020; Sköld-Chiriac et al., 2014) and increased predation risk (Laurila, Pilto & Hohtola, 2005; Carr & Lima, 2013; Andreasson, Nord & Nilsson, 2019).

(2) Morphological and anatomical adaptations

Feathers play a critical role in reducing thermogenic requirements by providing insulation, trapping pockets of warm air around the bird's body (Stettenheim, 2000). Birds keep their feathers clean by preening with oil from glands near the base of their tails to enhance their ability to trap air and thereby maintain their insulatory function (Ehrlich, Dobkin & Wheye, 1988; Sibley, 2020).

(a) Distribution of feathers

Feathers have multiple functions: they provide insulation (Whittow, 1986), protection against predation (Møller, Nielsen & Erritzøe, 2006), enable flight and contribute to signalling through their size, shape and colour (Foth & Rauhut, 2020; Andersson, 1982). Seasonal cold adaptation in birds includes changes in feather type, density and quality. The insulation properties of plumage depend on the thermal conductivity of the individual feathers and their collective ability to trap a layer of air (Lustick, 1984). An increase in feather mass and structural alterations has been documented in several birds during winter (Guallar & Jovani, 2020; Newton, 2009; Freed & Cann, 2012; Dawson, 2015). Contour feathers with downy lower barbs (‘after feathers’) are present in many birds only in the winter plumage (Osváth et al., 2018). The downy section is important for thermo-insulation and shows variability within species according to environmental conditions (Stettenheim, 2000; Gamero et al., 2015). For example, passerines (Passeriformes) in colder climates have longer feathers, with proportionately larger downy sections and lower downy-barb density than their counterparts inhabiting warmer regions (Wolf & Walsberg, 2000; Pap et al., 2017). Longer downy sections and more overlapping feathers increase the distance between the outside air and the bird's skin and increase the volume of trapped air, thereby increasing thermo-insulative capacity (de Zwaan, Greenwood & Martin, 2017; Stettenheim, 2000). An alternative to decreasing downy barb density to reduce heat loss is an increased pennaceous barb density that serves as a tighter seal against loss of warmed air (Butler, Rohwer & Speidel, 2008).

(b) Seasonal replacement of feathers

Adult birds of temperate regions generally moult and replace their feathers after the completion of breeding in late summer (Palmer, 1972). Small passerines such as the American goldfinch and house sparrow (Passer domesticus) moult into a winter plumage with an increased feather mass and number of barbules per length of barb (Freed & Cann, 2012). Alterations in feather texture and mass have also been observed in larger birds such as the common pheasant (Phasianus colchicus) and grey partridge (Perdix perdix) (De La Hera, Pérez-Tris & Tellería, 2010; Jenni & Winkler, 1994). There is variation in moult strategies: birds may undergo one annual moult, exhibit multiple moults each year, or may replace individual feathers according to insulative requirements – for example American goldfinches replace lost feathers selectively during cooler months to reduce total feather mass by the onset of the summer months (Dawson & Carey, 1976). Some birds also utilize changes in feathers during moulting to maintain camouflage against a changing background. For example, ptarmigan (Lagopus spp.), which inhabit tundra and cold alpine regions, go through multiple moults each year to match a background with variable snow cover in addition to altering their thermal insulation properties (Dawson, 2008; Coppack & Pulido, 2004).

(c) Other insulating strategies

Some birds (e.g. willow ptarmigan Lagopus lagopus, ducks and gulls) have thicker dermal scales to insulate their feet and legs to minimize heat loss (Swaddle & Witter, 1994; Duchamp & Barre, 1993; Stevens et al., 1986). Polar birds such as penguins, snowy owl (Bubo scandiacus) and willow ptarmigan protect themselves physically against the cold by growing a winter plumage or by relying on a layer of fat, somewhat similar to blubber (Lewden, Bonnet & Nord, 2020; Blix, 2016). Svalbard ptarmigan (Lagopus muta hyperborea) store up to 35% of their body mass as fat in preparation for winter (Stokkan, 1992). Adelie (Pygoscelis adeliae) and emperor penguins (Aptenodytes forsteri) have a plumage that provides exceptional insulation in air, but is compressed during diving, and thereby loses much of its insulating effect (Blix, 2016; Müller-Schwarze, 1984). During diving in ice-cold water, their subcutaneous fat layer provides essential insulation against heat loss. High-latitude ptarmigans have evolved plumage on their feet seasonally in the winter functioning as ‘snowshoes’ to reduce the energy costs of walking on snow (Stokkan, 1992).

(d) Vascular adaptations

Modification of the circulatory system is an integral component of cold adaptation in birds, functioning to conserve warmer blood in the body core. Many birds can limit heat loss from the blood flow to their legs via a counter-current heat-exchange mechanism (Goulden, 2016; Arad, Midtgård & Bernstein, 1989; Hillman & Scott, 1989). These intricate heat exchangers (tibio tarsal rete) have sophisticated vasomotor control activated by catecholamines (Farag, 2014; Reilly, Koelkebeck & Harrison, 1991; Arad et al., 1989). Areas of bare skin like combs and wattles are potential sites of rapid heat loss; to minimize this, their blood supply again passes through a dense multi-layered capillary network that functions as a counter-current heat exchanger (Midtgård, 1989; Robinson, Campbell & King, 1976). In addition, subcutaneous blood vessels can constrict, reducing the thermal gradient between the skin and environment, thereby reducing heat loss (Arad et al., 1989). Haematocrit [red blood cell (RBC) and haemoglobin content] and oxygen carrying capacity can both increase during cold adaptation to support the elevated oxygen demands of the thermogenic tissues; this is an energetically more economical strategy than relying on increased blood flow alone. In winter-acclimatized dark-eyed juncos, haematocrit increased by 11.1% and oxygen capacity by 8.6% relative to summer values (Swanson, 1990). Increased haematocrit during winter was also reported in American goldfinches (Carey & Morton, 1976) and white-crowned sparrows (Zonotrichia leucophrys) (deGraw, Kern & King, 1979).

III. SHIVERING IN BIRDS

In contrast to mammals, shivering has been documented both during acute and prolonged cold acclimation in birds (Hohtola, 2002, 2004). Shivering involves involuntary muscle contraction and relaxation that is activated by motor neurons and generates large amounts of heat.

(1) Molecular aspects of shivering

The pectoralis flight muscle of birds is highly enriched with mitochondria and serves as the major site of shivering thermogenesis (Yacoe & Dawson, 1983; Liknes & Swanson, 2011; Swanson et al., 2009; Bicudo, Bianco & Vianna, 2002; Petit & Vézina, 2014). In the house sparrow, increases in flight muscle mass, metabolic rate and oxidative metabolism have been associated with shivering (Zhang et al., 2015a; Swanson, Zhang & King, 2014). During shivering, several ATPases generate heat from the conversion of ATP, including myosin ATPase, Na+/K+-ATPase and sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) (Periasamy, Herrera & Reis, 2017; Bal et al., 2018; Nowack et al., 2017). Myosin ATPase hydrolyses ATP during the cross-bridge cycle, generating heat as a by-product (Rowland et al., 2015; Stewart et al., 2010; Cooke, 2011). Na+/K+-ATPase uses ATP to maintain ionic equilibrium across the sarcolemma, again producing heat (Nowack et al., 2017; Hohtola, 2002; Walter & Seebacher, 2009). Although, Na+/K+-ATPase has been proposed as a major contributor to obligatory thermogenesis, it is yet to be established whether it can be differentially regulated during shivering. SERCA is the primary Ca2+-uptake protein that lowers cytosolic Ca2+ levels, effecting muscle relaxation (Periasamy et al., 2011; Gamu et al., 2020). ATP utilization by SERCA has been suggested as a key contributor to shivering thermogenesis in avian skeletal muscle, which expresses SERCA abundantly (Nowack et al., 2017; Bicudo et al., 2001; Dumonteil, Barre & Meissner, 1995).

(2) Quality of shivering

During periods of inactivity birds utilize skeletal muscle tone (including postural activities) and/or shivering to produce heat. Neuronal control of shivering takes place both at the level of whole muscles and of individual motor units (Hohtola & Stevens, 1986). However, the exact correlation between episodes of tremor during shivering and heat production in the muscle has not yet been fully established.

Based on its intensity, shivering can be categorized as tremor or microvibrations. While, tremor is visible shivering, microvibrations can only be recorded with electromyograms (EMGs) (West, 1965; Tøien, 1992). At a molecular level, microvibrations may not be identical to visible shivering (tremor) but in some respects may resemble NST at least in terms of cytosolic Ca2+ transport. Data recorded from different endotherms indicate that not all muscle types can participate in shivering and that different muscles show distinct patterns of tremor and microvibrations (Carey, Johnston & Bekoff, 1989). Tremor frequency and intensity tend to reduce gradually during acclimation to a cold environment both in birds (Hohtola & Stevens, 1986) and mammals (Kleinebeckel & Klussmann, 1990). Microvibrations in EMG recordings from birds during cold acclimation indicate that shivering probably continues even after long-term exposure to cold (Hohtola & Stevens, 1986; Hohtola, 2004). Since thermogenesis may be required for hours or even days, the most aerobic, fatigue-resistant muscles and motor units are likely to be most suitable for shivering (Aulie & Tøien, 1988). The capacity for heat production due to shivering per unit muscle mass is presumably under selection and the efficiency of its chemo-mechanical coupling may vary to maximize survival (Scott, 2005; West, 1965). The capacity for shivering thermogenesis (microvibration/muscle tone) and NST may coexist in the skeletal muscle of birds, as both processes rely on high aerobic (mitochondria) capacity and high resistance to fatigue. A progressive increase in muscle tone with decreasing ambient temperature has been reported in rodents and may also apply to avian muscle (Bal et al., 2017). At a molecular level, increased muscle tone may involve similar ion (Ca2+) transport costs to those of NST (Grigg et al., 2022). Further, some portions of the avian pectoralis appear to be able to generate heat independent of motor or postural activity, perhaps suggesting that such parts are preferentially recruited for NST (Bicudo et al., 2001; Duchamp et al., 1999).

(3) Can shivering meet the thermogenic requirements of birds?

Based on the existing literature on mammals, we suggest that the thermogenic demands of birds may not be met solely by shivering thermogenesis (Nowack et al., 2019; Huo et al., 2022). Several lines of evidence suggest that NST is present in birds, at least at neonatal stages. Young birds with poor thermal insulation require greater heat production than can be generated by shivering (Dawson & Carey, 1976; Duchamp et al., 1999; Aschoff, 1981). Shivering (both number of episodes and intensity) of cold-acclimatized birds is significantly lower than that of warm-acclimated birds exposed to acute cold (Dawson & Marsh, 1989; Cui et al., 2019). After prolonged cold acclimation, two-month old male leghorns (young adults) and muscovy ducklings (Cairina moschata) do not shiver but maintain high oxygen consumption levels and euthermic Tb (El Halawani, Wilson & Burger, 1970; Barré et al., 1986a). However, a failure to document a correlation between EMG activity and the ability of birds to maintain Tb in the cold may not be sufficient proof of activation of NST (Marjoniemi & Hohtola, 2000; Hohtola, 2004). It has been proposed that increased muscle training and/or altered efficiency could both contribute to improved heat-production capacity of muscle during cold acclimation of birds (Swanson & Vézina, 2015; Rowland et al., 2015; Meizoso-Huesca et al., 2022; Li et al., 2021). However, this would also increase muscle tone, which may involve some molecular processes common to those activated in NST (see Table 1), calling for further research. Therefore, heat production arising from muscle training and tone may not only be derived from contraction and relaxation, but may also involve a NST component. A recent study comparing mouse and toad (Bufo bufo) muscle showed that resting muscle can generate heat in the SR without involvement of myosin ATPase-mediated activity, providing further evidence for muscle thermogenesis that does not involve shivering (Meizoso-Huesca et al., 2022).

IV. NST IN BIRDS

Several birds including chickens, ducklings and penguins are known to develop regulatory NST upon prolonged cold exposure (El Halawani et al., 1970; Duchamp et al., 1989b; Barré, 1984; Barre & Roussel, 1986), increasing their thermogenic capacity (Barre et al., 1985) and improving cold endurance (Dawson & Marsh, 1989; Duchamp et al., 1999). The site of avian NST remains a matter of debate, despite attempts since the 1920s to characterize the contributing organs. Cold acclimatization of house sparrows was shown to upregulate the metabolism of liver and skeletal muscle more than that of the brain (Barnett, 1968); the liver, kidneys and heart contribute about 60% of total heat production in birds (Clapham, 2012; Duchamp & Barre, 1993). NST in liver is thought to occur via mitochondrial ATP production and proton leakage (Brand et al., 2003; Else et al., 2004). Although early studies claimed to have identified BAT in birds, later work clarified that their winter fat is not a functional equivalent of classical mammalian BAT (Olson, Dawson & Camilliere, 1988; Saarela et al., 1989; Brigham & Trayhurn, 1994). Indeed, the mitochondria in this tissue have respiratory characteristics typical of those of mammalian WAT (Saarela et al., 1991). Other researchers argued that the multilocularity of adipocytes in winter fat of mammals and birds may indicate intense lipolytic activity to deliver fatty acids for thermogenesis in other tissues (Rabi, Cassuto & Gutman, 1977; Barré et al., 1986a). Several studies have investigated the role of skeletal muscle in avian NST and have characterized structural as well as functional changes during cold acclimation. The pectoral muscle of hummingbirds (~25% of body mass) with a mitochondrial density of 25–30% has been proposed as a candidate site for NST in addition to shivering, which also might be true for other avian species (Dawson & Carey, 1976; Dawson et al., 1983; Dawson & Marsh, 1989; Hart, 1962; Mathieu-Costello et al., 1998; Johnston et al., 1994; Eduardo, Bicudo & Chaui-Berlinck, 1998).

Many studies support the proposal that BAT-independent cold-induced NST is primarily of skeletal muscle origin in birds (Duchamp & Barre, 1993; Dumonteil et al., 1995). Skeletal muscle comprises more than 50% of body mass in birds [in Japanese quail (Coturnix japonica) it even reaches 70%], and thus is a strong candidate site for NST (Duchamp & Barre, 1993; Duchamp et al., 1991; Marjoniemi & Hohtola, 2000; Teulier et al., 2014). Birds that overwinter at cold latitudes undergo numerous biochemical changes in their muscles, especially in the flight muscles which comprise ~20% of body mass (Dawson & Carey, 1976; Dawson et al., 1983; Dawson & Marsh, 1989; Hart, 1962). These changes include improved lipid transport capacity and utilization (Marsh & Dawson, 1982; Zhang et al., 2015b). Blood flow and oxygen supply to the skeletal muscles remained high during long-term cold exposure of muscovy ducklings even in the absence of shivering as confirmed by simultaneous EMG recordings (Duchamp & Barre, 1993). Developmental studies using chicks, penguins and finches also provide support for skeletal muscle as a site of NST (Duchamp et al., 1991; Marjoniemi & Hohtola, 2000; Carey et al., 2019; Roussel et al., 2020). Precocial birds develop endothermy around the time of hatching, whereas altricial species do not do so until approximately 2 weeks after hatching. The ectothermy-to-endothermy transition in neonate birds is closely correlated with the development of skeletal muscle function and its oxidative capacity (Seebacher, Schwartz & Thompson, 2006; Lovegrove, 2017; Price & Dzialowski, 2018). Intriguingly, birds have a unique ability to utilize the same muscle for shivering and locomotion, unlike mammals. This represents a major impediment when designing experiments to segregate the molecular mechanisms contributing to locomotion, shivering and NST.

(1) Molecular mechanisms of muscle NST

There have been several attempts to isolate the molecular mechanisms underlying cold-induced NST in avian skeletal muscle (Teulier et al., 2014; Duchamp et al., 1991; Hirabayashi et al., 2005). Oxygen supply, free fatty acid load, and mitochondrial abundance all increase in muscle after cold acclimation, consistent with a role for muscle NST in addition to shivering (Stager, Swanson & Cheviron, 2015; Smith et al., 2018). However, the detailed molecular mechanisms remain debated.

(a) Na+/K+-ATPase activity and membrane leakiness

Skeletal muscles in birds adapted to cold maintain an elevated cellular ATP demand, proposed to be associated with activity of Na+/K+-ATPase involved in maintenance of ionic equilibrium across the sarcolemma (Turner et al., 2006; Else, Windmill & Markus, 1996). Although the exact mechanism is still unclear, it is thought that leakiness of Na+/K+ channels (which alter ion distributions across the cell membrane) can recruit ATP hydrolysis by Na+/K+-ATPases. The ATP-driven Na+/K+-ATPase is regulated by a key catalytic α1 subunit (Walter & Seebacher, 2009). A role of Na+/K+-ATPase activity and membrane leakiness in muscle NST was studied during the transition of hatchlings from ectothermy to endothermy (Hulbert & Else, 1990; Noble & Cocchi, 1990). An increase in membrane polyunsaturated fatty acids, which can elevate membrane leakiness, and increased messenger RNA (mRNA) levels of the catalytic α1 subunit were reported in chickens during this transition (Seebacher et al., 2006; Walter & Seebacher, 2009). Complete validation of muscle membrane leakiness and Na+/K+-ATPase activity involvement in NST awaits further research.

(b) Futile calcium cycling and free fatty acids

Several studies suggested that skeletal muscle-based NST in birds mainly relies on ATP-dependent Ca2+ cycling by SERCA which can be activated by Ca2+ leakage from the ryanodine receptor (RyR), possibly mediated by fatty acids. This cycling of Ca2+ between the cytosol and sarcoplasmic reticulum (SR) through SERCA without being coupled to activity of contractile filaments is termed ‘futile calcium cycling’. The suggestion that SERCA can function as a heat pump originated in studies performed on avian skeletal muscle (Duchamp & Barre, 1993; Block, 1994). Cold acclimation was shown to upregulate Ca2+-cycling proteins including SERCA and RyR in the skeletal muscle of ducklings, indicating an important role in muscle thermogenesis (Dumonteil, Barre & Meissner, 1993, 1995; Meissner & El-Hashem, 1992). Around 30–50% higher Ca2+ transport activity was observed in muscle homogenates and heavy SR microsomal fractions (laden with SERCA) from cold-acclimated ducklings (Dumonteil et al., 1995, 1993; Bicudo et al., 2001). A recent study showed that the proposed futile Ca2+ cycling operates in resting skeletal muscle of mice but not in muscle from toad (Rhinella marina) suggesting that minimal Ca2+ leakage from RyR1 can recruit SERCA-based ATP hydrolysis without inducing myosin ATPase activity (Meizoso-Huesca et al., 2022).

It has been suggested that the activity of SERCA working against the Ca2+ electrochemical gradient can undergo uncoupling, i.e. dissociation of Ca2+ uptake from ATP hydrolysis, leading to increased ATP utilization and the generation of heat (de Meis, Arruda & Carvalho, 2005; Kjelstrup et al., 2008; Marie & Silva, 1998). However, the mechanism by which SERCA uncoupling is regulated (i.e. whether simply by ionic concentration or by other routes) remains a matter of debate (Campbell & Dicke, 2018; Bal et al., 2018). The discovery of sarcolipin (SLN) as a SERCA uncoupler provides a potential regulatory mechanism. Although SLN function in muscle NST has been characterized in mammals (Bal et al., 2012, 2021; Oliver et al., 2019; Smith et al., 2002; Nowack et al., 2017, 2019; Wang et al., 2021), the role of avian SLN is still under-explored. In vitro and structural studies show that SLN is a single transmembrane protein that interacts with SERCA in the intramembranous region (Traaseth et al., 2008; Wang et al., 2021). SLN can bind to the Ca2+-bound form of SERCA and allow its ATP hydrolysis function while blocking Ca2+ transport across the SR membrane, thereby increasing ATP utilization and heat production. In mammals, cold acclimation has been shown to increase SLN expression in the skeletal muscle (Bal et al., 2012; Pant, Bal & Periasamy, 2015, 2016; Campbell & Dicke, 2018). Surprisingly, a study on adult dark-eyed juncos reported down-regulated SLN expression after cold acclimation (Stager & Cheviron, 2020). However, this study was performed using transcriptomic analysis without validation at the mRNA level or SERCA/SLN ratio determination and involved only extreme cold exposure. A recent study on migratory white-throated sparrows (Zonotrichia albicollis) showed that SLN is upregulated 25-fold during the period of migratory preparation, which might suggest a role in avian flight (Elowe & Gerson, 2022). However, recent molecular-level studies investigating how SLN interacts with SERCA demonstrated that SLN enhances SERCA-based heat production at least in mammalian muscle (Wang et al., 2021; Kaspari et al., 2020; Nowack et al., 2019). While any role of avian SLN awaits further study, it is possible that similar molecular mechanisms involving SLN–SERCA are present.

Activation of SERCA-mediated Ca2+ cycling in muscle in response to elevated cytosolic Ca2+ levels seems to be fundamental to NST, but how the cytosolic Ca2+ concentration can be elevated without initiating contraction is not fully explained. In mammals, RyR1-mediated SR Ca2+ release has been proposed as a mechanism (Sztretye et al., 2017), but involvement of other potential mechanisms such as Ca2+ entry through the sarcolemma via store-operated calcium entry (SOCE) has not yet been investigated. SOCE is mediated by stromal interaction molecule (STIM) and Orai proteins located on the SR and sarcolemma, respectively, that physically interact to move Ca2+ into the sarcoplasm from the extracellular fluid (Michelucci et al., 2018; Pan, Brotto & Ma, 2014; Edwards et al., 2010). Initially it was believed that SOCE is recruited only after depletion of the SR store, but later studies provided evidence that SOCE-mediated Ca2+ entry can act as a spatially localized signal to initiate specific Ca2+-signalling processes (Trebak et al., 2013; Launikonis, Murphy & Edwards, 2010). If SOCE can be regulated by an as-yet-undiscovered upstream mechanism during cold exposure that can activate SERCA and establish futile Ca2+ cycling without muscle contraction or shivering still requires clear demonstration.

A potential role of free fatty acids (FFAs), especially palmitoyl carnitine, in SR Ca2+ release has been investigated (Kochegarov, 2003; El-Hayek et al., 1993; Dumonteil, Barré & Meissner, 1994; Laver, 2007). In cold-acclimated muscovy duckling muscles, palmitoyl carnitine enhanced ryanodine binding in isolated SR and induced Ca2+ release from RyR1 (Dumonteil et al., 1994; Meissner, 1994). More recent studies have shown that RyR1-mediated Ca2+ release is involved in several pathophysiological conditions in mammals (Muslikhov, Sukhanova & Avdonin, 2014; Andersson et al., 2011; Eltit et al., 2010; Lamboley et al., 2016), where levels of muscle FFAs can be high. Levels of FFAs also are known to be elevated after cold acclimation in some birds and most mammals (Barré, Nedergaard & Cannon, 1986b; Freeman, 1967; Parker, 1978; Staples & Brown, 2008; Etches, John & Gibbins, 2008). Increased FFA uptake in the skeletal muscle with a concomitant rise in mitochondrial respiration was reported in cold-acclimated ducklings and pigeons (Barré et al., 1986b; Levachev et al., 1965; Roussel et al., 1998). Mitochondrial membrane potential measurements suggested that elevated ATP production in skeletal muscle supports ATP utilization by the Ca2+-cycling mechanism (Block, 1994; Kikusato & Toyomizu, 2013; Sommer et al., 2016; Ganitkevich, 2003). It seems plausible that FFAs both activate mitochondrial ATP synthesis and cause RyR1-mediated Ca2+ leakage, thus setting up futile Ca2+ cycling in the absence of muscle contraction. However, the mechanism of activation of futile Ca2+ cycling in avian skeletal muscle remains ambiguous and potential roles of endogenous agonists (e.g. thyroid hormones, glucagon, epinephrine) have been proposed.

(c) Mitochondrial inefficiency

Muscle NST has been proposed to be mediated by modulation of mitochondrial efficiency, via the coupling of proton gradient generation with ATP production. This could theoretically take place via several pathways: (1) leakiness of mitochondria; (2) uncoupling via the adenine nucleotide translocator (ANT); or (3) by action of avian UCP (avUCP), which was considered to be similar to mammalian UCP1.

(i) Leakiness of mitochondria

The leakiness of a biological membrane can be regulated by the composition of membrane lipids and the modulation (including expression) of ion channels. These modifications can influence the metabolic energy costs and the amount of heat production. At thermoneutrality, heat generated by such mechanisms can be considered as obligatory thermogenesis, but when this process is induced only during cold acclimation it should be considered as regulatory (facultative) thermogenesis. One mechanism proposed to explain avian muscle NST arises through inefficient mitochondrial metabolism. Any increase in the leakage of mitochondrial protons has been suggested to induce mitochondrial inefficiency and heat production. Oxidative phosphorylation required for ATP production is dependent on the presence of a proton gradient across the inner mitochondrial membrane generated by the mitochondrial electron transport system which consumes oxygen in the process (Salin et al., 2015). Leakage of protons across that membrane uncouples the link between oxygen consumption and ATP synthesis, reducing the efficiency of energy production but generating heat. The relative degree of coupling of the electron transport chain to ATP production may thus depend on the energetic requirements of the animal: proton leakage can be reduced to maximize ATP production during low substrate availability (Bourguignon et al., 2017; Glanville & Seebacher, 2010; Monternier et al., 2014; Trzcionka et al., 2008; Roussel et al., 2020). This has been studied in the mitochondria-rich BAT of mammals where such uncoupling is the basis for NST (Nicholls, 2013). However, analogous processes are reported for other tissues and organs, both in mammals (Blondin et al., 2017; Dawson et al., 2018; Laursen et al., 2015; Mahalingam et al., 2020) and in organisms that lack BAT (Trzcionka et al., 2008; Talbot et al., 2004; Toyomizu et al., 2002). While skeletal muscle mitochondrial leakiness has been suggested to occur in several mammals, the occurrence of a similar process in avian muscle has not been reported. Leaky mitochondria have been observed in avian RBCs, which do not lose their nucleus and organelles during maturation. Both mitochondrial volume and leakiness are upregulated in the RBCs of sympatric coal (Periparus ater) and great tit (Parus major) during the winter (Nord et al., 2021). However, any involvement of mitochondrial leakiness in avian muscle NST requires further investigation.

(ii) Uncoupling via adenine nucleotide translocator (ANT)

A mechanism of mitochondrial uncoupling involving ANT has been suggested based on its association with decreased mitochondrial membrane potential (Teulier et al., 2010). Upregulation of ANT expression at the mRNA level has been observed in avian skeletal muscle during cold acclimation (Toyomizu et al., 2002; Roussel et al., 2000; Mujahid et al., 2005; Ueda et al., 2005; Raimbault et al., 2001; Collin et al., 2003). Long-term exposure to cold increased ANT expression 1.7-fold in the subsarcolemmal mitochondria of male muscovy duckling gastrocnemius muscle (Roussel et al., 2000). Repeated exposure to cold water of juvenile king penguins (Aptenodytes patagonicus) led to enhanced expression and activity of avANT (Talbot et al., 2004). In the case of laying-type chickens, elevated expression of avANT mRNA was characterized during an adaptive response to mild cold with a concomitant increase in feeding efficiency (Mujahid et al., 2005). Upregulation of ANT is also observed in the skeletal muscle of cold-acclimated mammals, suggesting that ANT is part of a cold-mediated remodelling of skeletal muscle (Grigg et al., 2022; Kunji et al., 2016; Klingenberg, 2008; Bal et al., 2016). The major function of ANT is to catalyse the transmembrane exchange of cytosolic ADP with mitochondrial ATP across the inner mitochondrial membrane, serving as a key link between the mitochondrial and cytosolic compartments of cells. Therefore, it may not be feasible for ANT to act as an agent to physically uncouple the proton gradient (Talbot et al., 2004) and produce heat. The functional relevance of ANT upregulation might be to increase the rate of ATP delivery to the cytosolic side or regulation of the intra-mitochondrial adenine nucleotide pool size, thereby facilitating ATP-based NST.

(iii) Action of avian UCP (avUCP)

The expression of avUCP was increased in the skeletal muscle of cold-acclimated male muscovy ducklings (Rey et al., 2010), chickens (Toyomizu et al., 2002; Collin et al., 2003) and king penguins (Talbot et al., 2003, 2004), mRNA of avUCP was upregulated after long-term cold exposure in birds (Raimbault et al., 2001; Toyomizu et al., 2002; Collin et al., 2003), and its activity was increased after cold-water acclimation in penguins (Talbot et al., 2004). By contrast, levels of avUCP mRNA were unchanged following short-term cold exposure (Raimbault et al., 2001; Ueda et al., 2005; Mujahid et al., 2005), during which shivering is the main thermogenic mechanism. In cold-exposed chickens, avUCP expression in the skeletal muscle initially (1–2 days) increased significantly and then decreased slightly, but remained elevated even after ~10 days of cold acclimation (Ueda et al., 2005). Based on these findings, it was proposed that avUCP could form the basis of NST metabolic heat production in avian muscle analogous to the role of UCP1 in mammalian BAT (Raimbault et al., 2001; Collin et al., 2003; Vianna et al., 2001; Mujahid et al., 2005). However, subsequent studies showed that avUCP is not involved in heat generation and may not be the primary mediator of muscle NST in birds (Ueda et al., 2005; Criscuolo et al., 2006; Walter & Seebacher, 2009; Legendre & Davesne, 2020). However, avUCP is upregulated in fasting birds and thus might play a key role in FFA mobilization into the muscle (Collin et al., 2003; Evock-Clover et al., 2002; Abe et al., 2006; Toyomizu et al., 2006). A potential alternative role of avUCP in managing oxidative stress in skeletal muscle during cold acclimation also seems promising. Hence, further research is needed to clarify whether avUCP contributes to regulatory or obligatory thermogenesis.

(d) ROS metabolism

An elevated oxidative capacity in tissue is usually associated with increased reactive oxygen species (ROS) activity. Cold-adapted skeletal muscles are known to show an increased flux of lipid substrates and mitochondrial activity that might be associated with elevated ROS production in both mammals and birds. Interestingly, recent studies provide evidence that mitochondria are not the only source of ROS in muscle but that cytosolic ROS production by NADPH and xanthine oxidases are also equally important (Zorov, Juhaszova & Sollott, 2014; Di Meo, Napolitano & Venditti, 2019; Bouviere et al., 2021; Zhang & Wong, 2021). A mild increase in ROS levels helps in sustaining higher mitochondrial oxidative status during cold adaptation of avian skeletal muscle, while excess ROS production causes oxidative stress (Aksit et al., 2008). In a comparison of birds and mammals, it was shown that avian muscle has lower ROS levels than in equivalent-sized mammals, implying that birds are equipped with better ROS-handling mechanisms (Porter, Hulbert & Brand, 1996; Brand et al., 2003; Barja et al., 1994; Lambert et al., 2007; Buttemer, Abele & Costantini, 2010). Additionally, multiple lines of evidence suggest that birds have a much higher tolerance to ROS than do mammals. These include: a longer lifespan than mammals of similar body mass; their ability to withstand exposure to 95% oxygen for longer; and their 2–4 times higher blood glucose levels (Cohen et al., 2008). Other studies have demonstrated that upregulation of avUCP in cold-adapted skeletal muscle can modulate ROS metabolism and reduce its toxic effects, thereby serving as a facilitator of avian muscle NST (Talbot et al., 2004; Miwa & Brand, 2003; Rey et al., 2010). A higher ROS-handling ability of avian skeletal muscle might also have benefits such as enabling a longer dive capacity (e.g. in penguins) and better survival during cold adaptation (Corsolini et al., 2001). Interestingly, an increased plasma triiodothyronine (T3) level in ducklings is associated with down-regulated ROS production by the inter-myofibrillar mitochondria. This indicates a role of thyroid hormone (TH) in regulation of avian muscle metabolic rate that might contribute to overall NST (Rey et al., 2013).

V. HORMONAL REGULATORS OF COLD ADAPTATION IN BIRDS

As discussed above, effective cold adaptation in birds relies on both heat conservation and heat production mechanisms. These processes involve various organs and must be activated harmoniously. This coordination of multiple systems is orchestrated by the production of circulating hormones/neurohormones, of which the best understood in birds are thyroid hormone, catecholamines and glucagon (Fig. 1).

Details are in the caption following the image
Fig. 1
Open in figure viewerPowerPoint
Model depicting the major hormonal pathways that modulate non-shivering thermogenesis (NST) mechanisms in skeletal muscle. In birds, cold stress is proposed to affect levels of four circulating neurohormonal agents that influence molecular mechanisms associated with muscle NST. Glucagon and catecholamines may act as synergistic partners triggering muscle NST, while TH and corticosterone function together to reprogram the muscle during long-term cold adaptation. During cold exposure, increased influx of FFAs into the skeletal muscle via CD36 can initiate two distinct processes: mitochondrial ATP production and Ca2+ leakage from RyR1 leading to futile Ca2+ cycling by SERCA. Both glucagon and catecholamines facilitate uptake of FFAs into the muscle thereby activating muscle NST. Receptors for these hormones are located on the sarcolemma and are coupled with G-proteins. Binding of catecholamines (especially norepinephrine) to their receptors leads to production of cAMP which binds with calmodulin (a Ca2+-binding protein) to induce a cascade of intracellular events leading to enhanced mitochondrial activity. Glucagon is also suggested to boost mitochondrial ATP production in cold-exposed avian muscles although the mechanism is not completely defined. TH and corticosterones work in a coordinated manner to increase the expression pattern of nuclear thermogenic genes. Interestingly, the rise in the level of corticosterones prevents local conversion of T4 to T3 (the active form of TH) via Dio2. The effect of TH on increasing muscle oxidative capacity is primarily due to upregulation of mitochondrial content through several transcription factors like PGC-1α, PPARβ etc. TH has been shown to stimulate UCP3 and ANT expression that contribute towards increasing muscle oxygen consumption and resting metabolic rate. Corticosterones, especially glucocorticoids, work via several transcription factors, some of which are shared by TH and manipulate the process of mitochondrial biogenesis. AC, adenylate cyclase; ANT, adenine nucleotide translocator; Avβ3, alpha v beta 3; avUCP, avian UCP; B2AR, β2-adrenergic receptor; cAMP, cyclic adenosine monophosphate; CD36, cluster of differentiation 36; DHPR, dihydropyridine receptor; Dio2, type II iodothyronine deiodinase; FFAs, free fatty acids; FOXO1, forkhead box O1; GPCR, G-protein-coupled receptor; GR, glucocorticoid receptor; GRE, glucocorticoid response elements; Gs, stimulatory G protein; KLF15, Krüppel-like factor 15; OXPHOS, oxidative phosphorylation; PGC-1α, peroxisome proliferator-activated receptor γ coactivator 1α; PKA, protein kinase A; PPARβ, peroxisome proliferator-activated receptor β; REDD1, regulated in development and DNA damage responses 1; RyR1, ryanodine receptor 1; SERCA, sarcoendoplasmic reticulum calcium transport ATPase; SLN, sarcolipin; SNS, sympathetic nervous system; T3, triiodothyronine; T4, thyroxine; TH, thyroid hormone; TRE, thyroid response elements; TRH, thyrotropin-releasing hormone; TSH, thyroid-stimulating hormone; UCP3, uncoupling protein 3.

(1) Thyroid hormone

Thyroid hormone (TH) is thought to increase basal thermogenesis by lowering metabolic efficiency (Liu, Chen & Li, 2006) and facultative thermogenesis by facilitating tissue-level remodelling of the skeletal muscle for NST in birds (Collin et al., 2003; Decuypere et al., 2005; Yen, 2001; Meizoso-Huesca et al., 2022). The role of TH in cold adaptation of skeletal muscle has been studied extensively in several avian species. Plasma T3 (the physiologically active form) levels were higher during the cold season in red knots (Calidris canutus), tree sparrows (Passer montanus) and Chinese bulbuls (Pycnonotus sinensis) compared to their summer counterparts (Jenni-Eiermann, Jenni & Piersma, 2002; Zheng et al., 2014a; Zheng, Liu & Swanson, 2014b). Such elevated T3 levels increase both metabolic rate and cyclooxygenase (Cox, Winterbourn & Hampton, 2010) activity in muscle (Zheng et al., 2014b; Bahi et al., 2005; Irrcher et al., 2003; Cassar-Malek et al., 2007). Avian oxidative muscles are more sensitive to plasma TH levels than are their glycolytic muscles, in augmenting mitochondrial activity (Rey et al., 2013). In avian skeletal muscles, TH plays a critical role in developmental processes such as differentiation, fibre-type determination, mitochondrial biogenesis and oxidative capacity during the first weeks after hatching, especially in cold conditions (Dainat et al., 1986; Debonne et al., 2008; De Groef, Grommen & Darras, 2013; McNabb, 2006). Interestingly, induction of a hypothyroid state by methimazole decreases pectoralis muscle mass in domestic ducks, whereas induced hyperthyroid status (by TH administration) enhances muscle mass in chickens (Maruyama, Kanemaki & May, 1995; Bishop et al., 2000; Gardahaut et al., 1992; McNabb & Darras, 2015). Further, TH treatment promotes the production of slow fibres and enhanced aerobic capacity in locomotory muscles in young chickens (3–35 days of age) and adult ducks (McNabb & King, 1993; Dainat et al., 1991). TH regulates the expression pattern of proteins involved in muscle NST, such as Ca2+-handling proteins including SERCA, RyR1 and SLN; thereby adapting the framework of the skeletal muscle to perform as a site of NST (Bal & Periasamy, 2020; Grigg et al., 2022). Therefore, TH can be considered as a major moderator of long-term cold-induced acclimatization in avian skeletal muscle.

(2) Catecholamines

Studies in rodents showed that catecholamines (primarily norepinephrine) stimulate NST in BAT (Broeders, Bouvy & van Marken Lichtenbelt, 2015); while studies in monotremes and marsupials indicate that muscle NST can also be induced by catecholamines (Clements et al., 1998; Grigg, 2004; Jastroch, Polymeropoulos & Gaudry, 2021). In birds, thermogenic actions of catecholamines have been reported from studies both in vivo (Barre & Rouanet, 1983; Sutter & Macarthur, 1989; Hissa, Pyörnilä & Saarela, 1975a) and in vitro (Hissa, Rantala & Jeronen, 1975b). However, there have been contradictory findings regarding the role of catecholamines in thermogenesis during cold challenge in birds. While some have reported pronounced catecholamine-induced thermogenesis upon cold exposure in ducklings (Marmonier et al., 1997; Filali-Zegzouti et al., 2000; Slimani et al., 2007), king penguins and chicks (Barre & Rouanet, 1983); others found a much smaller effect than in rodents (Duchamp et al., 1999). Exogenous catecholamine injection during cold exposure can inhibit shivering in birds (Hissa, 1988), which might indicate activation of NST. Another study showed increased levels of plasma catecholamines in adult pigeons and chickens upon acute cold exposure within 30 min but decreasing to baseline levels within a few hours, while corticosterone levels remained elevated (Jeronen et al., 1976; Lin & Sturkie, 1968). Glucagon administration to ducklings led to a concomitant increase in plasma norepinephrine and metabolic rate, implying synergistic action of these two hormones (Filali-Zegzouti et al., 2005; Boomsma & Tipton, 2007).

(3) Glucagon

The effects of glucagon in birds are similar to those of catecholamines in mammals, leading to suggestions that glucagon could be a potential mediator of avian NST (Freeman, 1970; Barre, Cohen-Adad & Rouanet, 1987; Duchamp et al., 1999). Glucagon injection activates NST in birds like ducklings, penguins and pigeons (Barre et al., 1987; Duchamp et al., 1993, 1989a; Hohtola, Hissa & Saarela, 1977; Carey et al., 2019), but whether cold exposure increases plasma glucagon levels is not known. The physiological effects of glucagon might partly result from synergistic action with hormones that regulate metabolism and BMR, such as corticosterone and norepinephrine (Filali-Zegzouti et al., 2005; Koban & Feist, 1982; Abdelmelek et al., 2001). Glucagon has pronounced lipolytic effects in birds and causes FFA release from multilocular adipose tissues, leading to increased plasma FFA levels (Barré et al., 1986a; Bénistant et al., 1998). It was postulated that glucagon triggers muscle NST by either FFA-mediated mitochondrial uncoupling or FFA-induced RYR1 Ca2+ leakage, enhancing futile Ca2+ cycling via SERCA (Holz et al., 1999; Geisler, Kentch & Renquist, 2017; Dumonteil et al., 1995). These would result in increased muscle respiration and heat production upon activation in the cold. Further, in some experiments glucagon injection caused enhanced blood flow and oxygen content in the skeletal muscle in the absence of shivering (Duchamp et al., 1993; Montaron, Rouanet & Barré, 1995). Surprisingly, glucagon-induced thermogenesis is not observed in all avian species and may depend on age or acclimation status (Hohtola et al., 1977; Barré et al., 1986b; Duchamp et al., 1999). Moreover, it is unclear whether activation of muscle NST by glucagon results from direct action on myocytes or indirectly via increased lipolysis, which in turn would increase levels of FFAs leading to SR Ca2+ leakage.

VI. FUTURE DIRECTIONS

Different birds may have evolved varying strategies of NST to balance their thermogenic demands and the energy resources available. Although several mechanisms of avian NST have been proposed, the molecular details remain to be elucidated. Some key questions that future research should address are described below.

Gradual cold acclimatization leads to biochemical and ultrastructural changes in the skeletal muscles enabling them to function as a site of NST in addition to their locomotory and shivering functions (Fig. 2). Ca2+ has diverse roles in skeletal muscle; it initiates muscle contraction, determines ATP (energy) utilization and modulates gene expression patterns as a second messenger. Ca2+ levels in different ultrastuctural regions in the sarcoplasm of avian skeletal muscle are likely to undergo adaptive modifications during cold exposure. Characterization of these ultrastructural changes will be essential to understanding the importance of local Ca2+ levels in the process of cold adaptation. Mitochondria and SR are the Ca2+ storage sites in skeletal muscle, thus it is likely that ultrastructural alignment of the mitochondrial network with that of the SR and the T-tubular sarcolemma must be involved in cold adaptation. In mice, it has been shown that mitofusins (MFN1, MFN2) are upregulated in skeletal muscle during long-term cold exposure, indicating enhanced mitochondria–SR intimacy (Bal et al., 2016). However, whether such changes also occur in avian skeletal muscle after cold acclimatization has not been studied. Further, biochemical changes induced by Ca2+ signalling leading to alterations in gene expression patterns in cold-adapted avian skeletal muscle remain poorly defined. Recent human studies have indicated that muscle tone rather than visible shivering is important in the mediation of muscle NST (Periasamy et al., 2011; Meizoso-Huesca et al., 2022). Newer techniques such as fluorescence-labelled imaging, artificial-labelled electron microscopy, in-situ expression mapping, improved EMG recordings, etc., could be employed to understand the ultrastructural modifications that take place in avian skeletal muscle during cold adaptation.

Details are in the caption following the image
Fig. 2
Open in figure viewerPowerPoint
Summary of proposed major regulatory mechanisms of cold-induced remodelling of avian skeletal muscle. Where information is unavailable for birds, we use insights from research on mammalian species. Skeletal muscle undergoes concerted biochemical and ultrastructural remodelling during cold adaptation. Uptake of FFAs into muscle cells increases, which can induce two different pathways: (1) by entering mitochondrial oxidative metabolism leading to increased ATP production; and (2) by acting on RyR1 to cause Ca2+ leakage from the SR that leads to elevated cytoplasmic Ca2+ levels and recruitment of SERCA activity. FAUs such as CPT1, CPT2, and CACT facilitate entry of FFAs into the mitochondria. When cytosolic Ca2+ concentration reaches a threshold it activates myosin ATPase to establish cross-bridges with actin that forms the basis for heat production during shivering. SLN uncouples SERCA function, enhancing ATP utilization and increasing the overall energy demand of the muscle, simultaneously generating heat. While SERCA and RyR1 are upregulated in both mammals and birds, upregulation of SLN is reported only in mammals after cold acclimatization. Cytosolic Ca2+ also can activate master regulatory proteins like CamKII and NFAT through Ca2+-binding proteins like calmodulin and calcineurin. Several of these regulators act as TFs themselves or modulate other TFs manipulating expression of genes essential for cold adaptation. Activity of some of these TFs, including PGC-1α, PPARβ, NFAT and MEF2 upregulates mitochondrial biogenesis and induces mitochondria to undergo more fusion than fission, increasing their size and interconnectivity across the skeletal muscle, a key feature of cold adaptation. Sarcoplasmic Ca2+ concentration is influenced by several transporter proteins located in the sarcolemma, which can be grouped together as Ca2+ importers like NCX and Orai1. Ca2+ entry into the mitochondria can signal swift enhancement of ATP output during muscle NST activation. This can be regulated via two mechanisms: first, SR–mitochondria proximity is modulated by proteins like mitofusins (MFN-1 and MFN-2) that are upregulated during cold adaptation; second, VDAC–MCU located on the outer and inner mitochondrial membranes, respectively, are known to boost mitochondrial metabolism. ANT, adenine nucleotide translocator; CACT, carnitine-acylcarnitine translocase; CamKII, Ca2+/calmodulin-dependent kinase II; CD36, cluster of differentiation 36; CPT1, carnitine palmitoyltransferase; DHPR, dihydropyridine receptor; DRP1, dynamin-1-like protein GTPase; ETC, electron transport chain; FA, fatty acid; FAU, fatty acid uptaker; FFAs, free fatty acids; FIS1, mitochondrial fission 1 protein; MCU, mitochondrial calcium uniporter; MEF2, myocyte enhancer factor-2; MFF, mitochondrial fission factor; MFN, mitofusin; Mid 49/51, mitochondrial dynamics proteins of 49 and 51 kDa; NCX, sodium-calcium exchanger; NFAT, nuclear factor of activated T-cells; NST, non-shivering thermogenesis; OPA1, optic atrophy protein 1; Orai, calcium release-activated calcium channel protein; OXPHOS, oxidative phosphorylation; PGC-1α, peroxisome proliferator-activated receptor γ coactivator 1α; PPARβ, peroxisome proliferator-activated receptor β; RyR1, ryanodine receptor 1; SERCA, sarcoendoplasmic reticulum calcium transport ATPase; SLN, sarcolipin; SR, sarcoplasmic reticulum; TCA, tricarboxylic acid; TF, transcription factor; VDAC, voltage-dependent anion-selective channel.

Of the mechanisms proposed for avian skeletal muscle-based NST, SERCA-mediated processes seem the most promising. SERCA is important for contraction–relaxation activities, shivering and NST; hence its relative role in skeletal muscle NST has been difficult to dissect. In mammals, SLN-mediated uncoupling of SERCA resulting in heat generation has been identified as an important component in muscle NST (Bal et al., 2018). Interestingly, a SLN-like sequence is present in the genome of all avian species sequenced to date, suggesting an important role that requires its evolutionary conservation (Bal & Periasamy, 2020). However, the roles of SLN in avian NST have not been examined comprehensively (although see Stager & Cheviron, 2020). As extreme cold exposure induces all heat-generating mechanisms, including shivering, any role of SLN in NST might be concealed. Hence, future studies should focus on adaptation to moderate cold in birds, where shivering recruitment is low. Further, SLN should be studied in different clades of birds and different muscle groups to throw light on the evolutionary/adaptive significance of SLN-based muscle NST. From studies in mice it is clear that SLN functions are most effective in postural muscles (Sopariwala et al., 2015), avian SLN also might be recruited differentially for thermogenesis in particular muscle groups. Future studies should validate RNA-sequencing data by utilizing gene-specific methods such as real-time polymerase chain reaction (RT-PCR), western blotting, and immunohistochemistry. The SERCA to SLN ratio is a major determinant of SLN function in mammals, and should be carefully evaluated in different muscle groups in birds. SR membrane lipid composition also has been suggested as a factor affecting SLN binding to SERCA in mammals (Funai et al., 2013) but has not been investigated in birds during cold acclimatization.

Avian NST might be differentially recruited depending on the severity of cold challenge, allowing improved survival while minimizing energy cost. Such differential NST activation is suggested to be regulated in birds by four hormones; TH, glucagon, catecholamines, and corticosterone. The exact roles of these hormones in shivering versus NST and activation versus programming has not been fully defined. Future investigations should measure plasma concentrations of these hormones during acute cold exposure and after long-term cold acclimatization. TH, glucagon and epinephrine have been proposed to regulate Ca2+-cycling-mediated muscle NST in response to cold (Bal & Periasamy, 2020). Glucagon either alone or in combination with norepinephrine has been shown to increase circulating levels of FFAs in birds leading to heat production in the skeletal muscle (Grande & Prigge, 1970; Barre & Rouanet, 1983). The involvement of TH in avian NST is likely to facilitate long-term adaptive changes in the thermogenic tissues via control of gene expression (McNabb & Darras, 2015). Corticosterone might have similar effects as TH on avian skeletal muscle, although some studies have suggested a role in acute cold challenge (Ruuskanen, Hsu & Nord, 2021). There are inconsistent data on the effects of corticosterone on resting metabolic rate (RMR) in birds, with some studies reporting a positive correlation (Jimeno et al., 2020; Jimeno, Hau & Verhulst, 2017) but others showing no correlation (Buttemer, Astheimer & Wingfield, 1991). Another interesting question that remains unanswered is how hormone–behaviour inter-relationships help birds to adapt to climatic seasonality, and thereby to long-term cold adaptation.

Skeletal muscle of cold-adapted birds exhibits increased oxidative metabolism, regulated ROS activity and upregulation of avUCP and ANT. While initial studies proposed that avUCP and ANT function like mammalian UCP1 in mitochondrial uncoupling to generate heat, later studies refuted this idea (Raimbault et al., 2001). Our current understanding is that avUCP works as a modulator of ROS production in skeletal muscle mitochondria. This antiradical function of avUCP would be of particular importance in preventing cellular damage during the hyper-metabolism that occurs during chronic exposure to cold (Walter & Seebacher, 2009; Rey et al., 2010). But whether skeletal muscle can simultaneously utilize avUCP-mediated heat generation via mitochondrial uncoupling and higher mitochondrial ATP production via oxidative phosphorylation still awaits clear demonstration. Studies investigating the rescue of UCP1-KO from cold by skeletal muscle-specific over-expression of avUCP or investigating cold adaptation of avUCP-KO birds could provide much-needed insights.

Most of the evidence for muscle NST in birds come from studies on hatchlings or juveniles. Currently, it is generally thought that in adult birds with fully developed skeletal muscle shivering predominates and NST may not be part of acclimation to cold. Much of the published literature on cold acclimation of adult birds has examined changes in hormones and related parameters, but molecular aspects have not been thoroughly investigated. Studies of structural and biochemical adjustments in the skeletal muscle after cold acclimation in adult birds would be illuminating.

VII. CONCLUSIONS

(1) Our review highlights that birds (at least young birds) are unlikely to be able to maintain their Tb in cold climates by heat conservation and shivering, and therefore presumably recruit NST. Understanding avian thermogenesis could provide important insights regarding the mechanisms involved in BAT-independent NST and potentially advance our understanding of mammalian NST. In general, NST is employed at temperatures below thermoneutrality and its extent should be tuned to the severity of cold. Although the exact mechanisms of avian NST are currently poorly defined, several conclusions are possible. Birds have evolved a series of organ- and whole-body-level adaptations to enable them to cope with seasonal cold challenges. These adaptations allow them to minimize heat loss and conserve heat in the core of the body thereby helping in Tb maintenance. However, these mechanisms may not be developed fully in young birds and may be insufficient during acute (or diurnal) cold. In such conditions, facultative thermogenesis including shivering and NST are likely to play important roles.

(2) Shivering is an acute defence mechanism that provides time for the activation of other mechanisms of thermogenesis including NST. In shivering, myosin ATPase serves as the prime mechanism of heat production. But shivering alone may provide insufficient heat in young birds and when there is a sudden large drop in ambient temperature. Research primarily from young birds provides evidence for skeletal muscle as the site of NST. Whether skeletal muscles of adult birds employ NST or only shivering is currently not fully defined.

(3) Several potential molecular mechanisms have been proposed as basis for avian skeletal muscle NST, of which SR-induced SERCA-based ATP utilization is the most promising. Interestingly, SLN acts as a SERCA regulator in mammalian skeletal muscle, but studies in birds are still rudimentary. The other potential mechanism of muscle NST in birds is ANT-based futile mitochondrial energy dissipation.

(4) It seems clear that avUCP is not a functional homolog of mammalian UCP1 even though it shares some sequence similarity. It is becoming evident that avUCP functions in the prevention of mitochondrial ROS production and oxidative stress in avian skeletal muscle under situations of intense metabolic activity, hence indirectly serving as a facilitator of NST.

(5) Glucagon has been suggested by some studies as a trigger for avian muscle NST. While TH serves to elevate muscle NST during long-term adaptation to seasonal cold, catecolamines seem to be more important during acute activation of muscle NST.

ACKNOWLEDGEMENTS

P.P. is highly indebted to the Government of India, Ministry of Science & Technology, Department of Science & Technology for pre-doctoral fellowship (DST/INSPIRE Fellowship/2018/IF180892). N.C.B. is funded by the Science and Engineering Research Board (SERB), DST, India (ECR/2016/001247) and DBT, India (BT/RLF/Re-entry/41/2014 and BT/PR28935/MED/30/2035/2018). We thank Professors Madhab Chandra Dash and Muthu Periasamy, Mr. Sunil Pani and the two anonymous reviewers for constructive critical suggestions and discussions that helped us to shape this review.

      The full text of this article hosted at iucr.org is unavailable due to technical difficulties.