Sarcolipin relates to fattening, but not sarco/endoplasmic reticulum Ca²⁺-ATPase uncoupling, in captive migratory gray catbirds Free

Migratory songbirds undergo physically demanding journeys each year between breeding and non-breeding locations. They accomplish this through a series of long-duration, non-stop flapping flights operating at metabolic rates up to 18 times their basal metabolic rates (Jenni-Eiermann, 2017). This physiological demand is met by the anticipatory development of a ‘migratory syndrome’ of traits that enable successful migration through a suite of physiological changes (Dingle, 2006), including adaptive changes to metabolism and body composition that are particularly evident in the increased fat stores and capacity to transport and catabolize fat as a fuel (Guglielmo, 2010; McFarlan et al., 2009; Ramenofsky et al., 2017; Zajac et al., 2011).

To accommodate extended periods of flapping flight, bird muscle undergoes hypertrophy and changes to efficiency in preparation for migration. Alterations to the myosin heavy chain (Velten et al., 2016), fiber hypertrophy (DeMoranville et al., 2019; Marsh, 1984), capillary density (Lundgren and Kiessling, 1988) and corticosterone signaling (Pradhan et al., 2019) have all been documented in flight muscles of migratory birds. One important source of energy expenditure in the muscle is Ca²⁺ sequestration in the lumen of the sarcoplasmic reticulum (SR) against a considerable concentration gradient. This is accomplished by the sarco/endoplasmic reticulum Ca²⁺ ATPase (SERCA), which under ideal conditions pumps Ca²⁺ into the SR at a coupling ratio of 2 Ca²⁺ ions for every 1 ATP hydrolyzed (Toyoshima, 2008). The maintenance of this Ca²⁺ gradient by SERCA can account for as much as half of the resting metabolic rate of mouse skeletal muscle (Smith et al., 2013).

The high energetic cost of maintaining the Ca²⁺ gradient may be exacerbated by increased Ca²⁺ leakiness of the SR membrane or by the addition of uncoupling proteins such as sarcolipin (SLN). SLN binds to SERCA and uncouples Ca²⁺ transport from ATP hydrolysis, increasing the duration of Ca²⁺ transients in the cytoplasm and futile cycling of SERCA (Smith et al., 2002). In rodent models, this leads to heat production (Bal et al., 2018; Nowack et al., 2017; Periasamy et al., 2017), slower muscle relaxation (Tupling et al., 2011), higher resting energy expenditure (Bombardier et al., 2013) and reduced obesity (Maurya et al., 2015). Recent studies also demonstrate SLN-mediated increases in cytosolic Ca²⁺ signaling for greater peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α) expression, leading to mitochondrial biogenesis, fatty acid catabolism and greater fatigue resistance in skeletal muscle (Fajardo et al., 2017; Maurya et al., 2018; Sopariwala et al., 2015; Summermatter et al., 2011). PGC1α also plays an important role in mediating muscle plasticity in response to exercise, in part due to alterations in Ca²⁺ signaling and energy use (Hoppeler, 2016; Summermatter et al., 2011). Furthermore, in Sln-knockout mouse studies, the loss of SLN expression leads to impaired expression of multiple genes including carnitine palmitoyltransferase 1 (CPT1), long-chain acyl-CoA dehydrogenase (LCAD) and 3-hydroxyacyl-CoA dehydrogenase (HOAD), as well as reduced fatty acid oxidation, all of which are associated with the migratory phenotype of songbirds (Marsh, 1981; McFarlan et al., 2009; Price et al., 2010; Sharma and Kumar, 2019; Zhang et al., 2015b).

We recently found evidence of approximately 25-fold increased SLN expression and unchanged SERCA2 transcript levels alongside ∼2-fold higher PGC1α mRNA in the pectoralis flight muscle of white-throated sparrows (Zonotrichia albicollis) in the migratory condition (Elowe and Gerson, 2022). While the uncoupling action of SLN could logically impede efficient use of ATP and slow rates of muscle contraction during long-distance migratory flight, Ca²⁺ signaling in the muscle may also facilitate the shift towards fat metabolism and mitochondrial biogenesis in the flight muscle of migratory birds. Therefore, SLN uncoupling may act as a signaling mechanism for a suite of changes in muscle physiology during spring migration. Alterations to Ca²⁺ dynamics in the muscle may appear through changes to SERCA activity or Ca²⁺ uptake rates, or the apparent coupling ratio of the two measurements as an indicator of SERCA efficiency (Bombardier et al., 2013; Fajardo et al., 2015). However, Ca²⁺ dynamics can also vary with SR membrane Ca²⁺ leak, leading to lower SERCA efficiency without the involvement of uncoupling proteins directly (Fajardo et al., 2015).

To investigate the role of SLN in the development of the migratory phenotype, we conducted a photoperiod manipulation using gray catbirds (Dumetella carolinensis) and monitored body composition in their transition to the migratory condition. Numerous studies have applied photoperiod manipulations to stimulate a transition into the migratory condition in birds, revealing physiological and behavioral adaptations for migration (Gwinner, 1990; Owen et al., 2006; Ramenofsky et al., 2003; Zajac et al., 2011). We tracked body composition changes from a short day ‘winter’ photoperiod (8 h light:16 h dark) to a long-day ‘spring’ photoperiod (16 h light:8 h dark) and monitored the development of migratory restlessness overnight, or Zugunruhe in order to confirm migratory condition. After 3 weeks and consistent overnight Zugunruhe, we measured pectoralis flight muscle mass and collected tissue to measure transcription of genes involved in the regulation of Ca²⁺ cycling and signaling (SLN, SERCA2 and PGC1α) and multiple indicators of the migratory phenotype, including fatty acid binding (heart-type fatty acid binding protein, h-FABP), fat transport (fatty acid translocase, CD36) and β-oxidation (HOAD). We also measured SERCA activity and Ca²⁺ uptake rates in the muscle to explore changes in SR membrane Ca²⁺ leak and SERCA efficiency and mitochondrial biogenesis using relative mitochondrial DNA (mtDNA) copy number. We hypothesized that gray catbirds increase SLN expression in the migratory condition, altering SERCA efficiency and signaling for seasonal changes to PGC1α, fat metabolism, body composition and mitochondrial biogenesis.

Gray catbirds, Dumetella carolinensis (Linnaeus 1766) (‘catbirds’ hereafter) are 30–50 g passerines that breed in North America from Nova Scotia to eastern Washington state and into the southern reaches of Alberta and British Columbia. They migrate south in the winter to the Gulf of Mexico, with some birds remaining in the southern USA and others to the Yucatán Peninsula down to Panama, with a small number of birds lingering along the eastern seaboard of the USA up to New York. Band recaptures suggest that birds from the northeastern USA migrate to Florida and the Caribbean islands (Ryder et al., 2011). We captured 20 catbirds (10 male, 10 female) by mist-netting during autumn migration between 29 September and 10 October 2019 on the University of Massachusetts Amherst campus (42°23′45.9"N 72°31′04.5″W). Average capture mass was 41.17±3.06 g. We immediately transported them to captive facilities at the University of Massachusetts Amherst where they were initially housed with 2 individuals per cage (77.5×30.5×39 cm) at 21°C and exposed to a decreasing autumn photoperiod to match the duration of civil dawn to civil twilight, with full-spectrum lights adjusted every 2–3 days until a short-day winter photoperiod (8 h light:16 h dark) was achieved. A dim light at night provided ∼1 lx in the room (Ramenofsky et al., 2003). Birds had ad libitum access to water and a synthetic high-carbohydrate diet modified from Guglielmo et al. (2017) with supplemental Tenebrio mealworms provided daily (Dick and Guglielmo, 2019). They were kept on the short-day winter photoperiod for 70 days to break photorefractoriness (Barceló et al., 2016). Our study followed the Institutional Animal Care and Use Committee guidelines approved by the University of Massachusetts Amherst (protocol 2018-0038). A collection permit was granted by the US Fish and Wildlife Service (permit MB65968B-1) and the State of Massachusetts (192.19SCB) to A.R.G.

Sex was determined by molecular sexing using whole blood prior to the photoperiod manipulation (Griffiths et al., 1998). We randomly divided half of the catbirds, with equal numbers of males and females, into two adjacent rooms where they were singly housed (38.5×30.5×39 cm) and allowed to acclimate for 19 days to the new space. On 24 March 2020, one room was switched to a long day spring photoperiod (16 h light:8 h dark) to induce a migratory disposition (Gwinner, 1990; Owen et al., 2006; Ramenofsky et al., 2003) while the other room remained on the same short-day winter light regime. Infrared cameras (Ailipu Technology Co., Ltd, Guangdong, China) were used to monitor night-time activity in both rooms following the change in light cycle. Catbirds held on the long-day light regime (N=5 males, N =5 females; ‘LD’ treatment birds, hereafter) showed substantial variation in their timing for displaying migratory restlessness, with one bird starting the first night and the next beginning on day 7 following the light change, with all birds except one showing migratory restlessness by day 14. Males and females did not appear to show a difference in the development of Zugunruhe. Catbirds kept on the short-day photoperiod (N=5 males, N=5 females; ‘SD’ treatment birds, hereafter) displayed little to no nocturnal activity.

During the photorefractory period, we weighed birds on three separate occasions and inspected for mass changes, health, and muscle and fat scores. Starting on the day of the photoperiod switch, we tracked changes in body composition using a quantitative magnetic resonance body composition analyzer (QMR) customized for small birds (Echo-MRI Echo-Medical Systems, Houston, TX, USA). The QMR measures fat and wet lean masses with accuracies of ±6–11%, and ±1–2%, respectively (Guglielmo et al., 2011). To ensure scans were always of post-absorptive birds, food was removed from cages approximately 2 h prior to scanning. Birds were scanned on day 1, 7 and 14, and immediately prior to euthanasia by isoflurane overdose and cervical dislocation (in compliance with all animal care guidelines), 21–22 days after the photoperiod switch, between 10:30 h and 16:30 h. Pectoralis samples were frozen on dry ice within 5 min of euthanasia for gene expression. We weighed the left pectoralis to the nearest 0.0001 g and dried small samples in an oven at 60°C for ∼48 h to estimate organ dry mass.

We homogenized approximately 50 mg of pectoralis muscle using beadmill homogenization (NextAdvance, Troy, NY, USA; speed 8, time 4, 4°C) in 1 ml of TriZol (Invitrogen, Carlsbad, CA, USA) and RNA was separated into an aqueous phase using a chloroform–ethanol procedure. Total RNA was purified and DNase treated using a Monarch RNA Cleanup Kit (New England Biolabs, Ipswich, MA, USA). Samples were quantified and checked for RNA quality using a Nanodrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA) prior to reverse transcription to cDNA using Luna RT (New England Biolabs). Each cDNA sample was diluted 1:5 with nuclease-free water before analysis.

We designed primers for gray catbirds using conserved sequence alignments. We began by using NCBI Blast (Madden, 2003) for each gene transcript against Gallus gallus and the Passeriformes taxonomy to obtain transcript sequence alignments across diverse songbirds. Using PrimerIdent (Pessoa et al., 2010), we selected candidate primers for target genes based on conserved locations in the transcript sequences. Our target genes were SLN, SERCA2, CD36, h-FABP, HOAD and PGC1α. For housekeeping genes, we selected hypoxanthine phosphoribosyltransferase 1 (HPRT1) as a stably expressed gene in multiple tissues in rats (Kim et al., 2014) and ribosomal protein lateral stalk subunit P0 (RPLP0), which did not show seasonal changes in gray catbirds previously (DeMoranville et al., 2019) and has been evaluated as a reference gene in several bird species (Olias et al., 2014) (see Table 1 for primer information). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) is frequently used as a housekeeping gene, but we previously found minor differences in expression under photoperiod manipulations (Elowe and Gerson, 2022) that make this gene a less reliable option.

Prior to running samples, we confirmed primer specificity on a 2% agarose gel in TAE with SYBR-safe stain and tested efficiency on a serial dilution series of pooled cDNA (1:2 to 1:40). All amplification efficiencies fell between 97.9% and 99.7%. We prepared reactions on ice in 10 μl volumes with 0.2 μl primers and 1 μl cDNA in Luna Universal qPCR master mix (New England Biolabs) to run on a fast protocol in a StepOnePlus qPCR machine (Applied Biosystems, Foster City, CA, USA) for 60 s at 95°C, then 40 cycles of 95°C for 15 s followed by 60°C for 30 s, with melt curve analysis from 60 to 95°C. We performed comparative cycle threshold (C_t) analysis according to Schmittgen and Livak (2008). Each sample was run in duplicate wells and we subtracted average C_t values for each target gene (SLN, SERCA2, CD36, h-FABP, HOAD and PGC1α) from the average C_t for the two corresponding housekeeping gene readings (HPRT1 and RPLP0) to obtain the ΔC_t for each individual and gene. Both HPRT1 and RPLP0 were stably expressed between the photoperiod treatments (HPRT1: t_14.8=0.55, P=0.589; RPLP0: t_12.8=−1.05, P=0.314; average: t_12.6=−0.25, P=0.808) and sex (HPRT1: t_14.8=0.55, P=0.589; RPLP0: t_12.7=−1.47, P=0.165; average: t_13.7=−1.30, P=0.214). Fold-change calculations incorporated the mean and s.e.m. for ΔC_t by subtracting the SD mean ΔC_t from the LD mean to provide ΔΔC_t, from which we calculated mean fold-change and upper and lower limits of the s.e.m. using 2^−(ΔΔC_t).

We tested for relative mitochondrial DNA copy number using qPCR comparing C_t for mitochondrial DNA (mtDNA) to genomic DNA (gDNA) (Castellani et al., 2020; Maurya et al., 2018). We extracted and purified DNA from frozen pectoralis muscle tissue using Monarch gDNA Purification Kits with Tissue Lysis buffer (New England Biolabs). Samples were quantified and checked for DNA quality using a Nanodrop 2000 (Thermo Fisher Scientific) and diluted to 15 ng μl⁻¹ prior to qPCR. We designed primers for mitochondrial DNA around the mitochondrial 16S rRNA gene and for genomic DNA around the BECN1 gene using conserved sequence alignments for Passeriformes (Table 1) and confirmed primer specificity as above. Amplification efficiencies were between 96.2% and 97.5% for mtDNA and gDNA, respectively. We prepared reactions on ice in 10 μl volumes with 0.25 μmol l⁻¹ primers and 1.5 ng μl⁻¹ DNA in Luna Universal qPCR master mix (New England Biolabs) to run in a Stratagene MX3000P qPCR system (Agilent, Santa Clara, CA, USA) for 60 s at 95°C, then 40 cycles of 95°C for 15 s followed by 60°C for 30 s, with melt curve analysis from 60 to 95°C. Relative mtDNA copy number was calculated by subtracting the C_t for gDNA from that for mtDNA (ΔC_t).

We measured total SERCA activity using an enzyme-linked spectrophotometric method as described elsewhere (Fajardo et al., 2015; Gamu et al., 2019; Jannas-Vela et al., 2020; Tupling and Green, 2002). Frozen pectoralis muscle tissue (∼80 mg) was homogenized in 1:10 ice-cold buffer (in mmol l⁻¹: 250 Sucrose, 5 Hepes, 0.2 PMSF, 15 NaN₃, pH 7.5) in a ground-glass conical handheld homogenizer. Aliquots of ∼150 µl of the homogenate were immediately stored at −80°C until use and never exceeded three freeze/thaw cycles. The SERCA ATPase reaction buffer contained (in mmol l⁻¹) 200 KCl, 20 Hepes, 15 MgCl₂, 10 phosphoenolpyruvate, 1 EGTA, 10 NaN₃ (pH 7.0 at 39°C). Prior to starting the reaction, we added 18 U ml⁻¹ lactate dehydrogenase (LDH), 18 U ml⁻¹ pyruvate kinase (PK), 5 mmol l⁻¹ ATP, 5 μmol l⁻¹ CaCl₂ (corresponding to free [Ca²⁺] of approximately 1000 nmol l⁻¹, pCa=6) and 0.3 mmol l⁻¹ NADH. Samples were run with and without 4 μmol l⁻¹ calcium ionophore A23187 (Sigma-Aldrich C75221), which allows Ca²⁺ leak through the SR membrane and prevents back-inhibition, or 50 μmol l⁻¹ SERCA inhibitor cyclopiazonic acid (CPA; Sigma-Aldrich C1530) to selectively inhibit SERCA enzyme activity. We ran 150× diluted homogenate in SERCA ATPase reaction buffer in duplicate in a clear 96-well plate with 200 μl reaction volume to measure the conversion of NADH to NAD⁺ using 340 nm absorbance on a Biotek Synergy H1 microplate reader (Agilent) for 40 min at 39°C, a typical avian body temperature. We measured SERCA activity with and without the Ca²⁺ ionophore, corrected for background activity in the presence of the SERCA inhibitor CPA and normalized activity to homogenate protein content measured using a Bradford reagent assay (Thermo Scientific 23200). We estimated membrane Ca²⁺ leak through the SR membrane by comparing maximal rates with and without the Ca²⁺ ionophore, with ratios closer to 1 indicating greater SR membrane permeability to Ca²⁺ (Fajardo et al., 2015).

All statistics were performed in R (v4.2.1, http://www.R-project.org/). To account for body size, we included sternum measurements in each model. Sternum size was significantly related to lean mass (t₁₈=2.37, P=0.029, R²=0.20), but not fat mass (t₁₈=1.30, P=0.210, R²=0.04), making it a suitable proxy for the structural body size of an individual. For body composition changes over time, we evaluated linear mixed models with the function ‘lmer’ (lme4 package, v1.1-17). We evaluated body composition (total, fat or lean mass) with terms for time point×photoperiod with sternum size as fixed effects while controlling for repeated measures in individual birds with a bird ID random effect. For SLN and SERCA2 ΔC_t gene expression data, we started with an ANCOVA with the photoperiod treatment and sternum size to account for structural body size. To evaluate the relationship between SLN expression and physiological measurements regardless of photoperiod treatment group, we used linear models with predictors for SLN ΔC_t gene transcription and sternum size to account for structural body size to test for correlations with ΔC_t data for other genes and pectoralis mass, SERCA activity, Ca²⁺ uptake, ionophore ratio, apparent coupling ratio, and fat and lean mass. For all models, we evaluated for normal distribution and equal variance of residuals.

At the start of the experiment, there was no significant difference in total body, fat or lean mass between the treatment groups (P>0.1; starting body mass 38.17±3.34 g; fat 4.30±2.72 g; lean 29.06±1.69 g; Fig. 1). Relative to the start of the photoperiod switch, we found a significant time×photoperiod interaction (F_3,60=5.12, P=0.003), with LD birds showing 3.50±1.04 g greater total body mass by day 14 after the photoperiod switch. We also found a significantly greater body mass with larger sternum size (F_1,20=6.32, P=0.021). We found a significant time×photoperiod interaction in fat mass (F_3,60=9.14, P<0.001), with LD birds showing 4.05±0.97 g greater fat mass by day 14 after the photoperiod switch than SD birds. Fat mass was not related to sternum size (F_1,20=0.45, P=0.51). For lean mass, we found a significant time×photoperiod interaction (F_3,60=4.60, P=0.006), with LD birds showing 1.03±0.38 g less total lean mass than SD birds by the end of the experiment. Lean mass was strongly correlated with sternum size as well (F_1,20=10.74, P=0.004).

We found approximately 6-fold higher SLN transcription in the LD photoperiod treatment birds relative to the SD birds (F_1,17=11.43, P=0.004) with no corresponding differences in SERCA2 transcripts (F_1,17=0.28, P=0.603; Fig. 2). Across the other genes, SLN ΔC_t showed a significant positive correlation with h-FABP transcription (t₁₇=8.04, P<0.001, R²=0.78) and PGC1α (t₁₇=2.26, P=0.037, R²=0.17), but no relationship with SERCA2, CD36 or HOAD transcription (P>0.1; Fig. 3).

SLN transcription (ΔC_t) was very strongly correlated with total fat mass (t₁₇=−7.01, P<0.001, R²=0.71), but not lean mass (t₁₇=1.66, P=0.12). SLN transcription was not correlated with pectoralis mass (wet: t₁₇=0.74, P=0.467; dry: t₁₇=1.66, P=0.11; Fig. 4).

Overall, we found no relationships between SLN transcription and measurements of SERCA activity without the calcium ionophore (t₁₇=1.18, P=0.292; Fig. 5A) or with the ionophore (t₁₇=0.81, P=0.431; Fig. 5B) or the ratio of the two measures (t₁₇=0.57, P=0.460; Fig. 5C). Similarly, there was no significant relationship with Ca²⁺ uptake rate (t₁₇=1.19, P=0.249; Fig. 5D) or the apparent coupling ratio of SERCA activity to Ca²⁺ uptake (t₁₇=1.13, P=0.273; Fig. 5E). Finally, we found no significant relationship between SLN and the ΔC_t between mtDNA and gDNA (t₁₇=−0.10, P=0.92; Fig. 5F).

Despite the lack of a relationship with SLN transcription, we found a non-significant trend between SERCA coupling ratio and mtDNA copy number (t₁₇=−1.72, P=0.104; Fig. 6A) across treatments, with lower SERCA coupling associated with greater mtDNA. This appears to be driven primarily by a significant correlation with the rate of Ca²⁺ uptake (t₁₇=−2.26, P=0.037; Fig. 6B) rather than the maximal SERCA activity (t₁₇=−1.60, P=0.128; Fig. 6C), with lower rates of Ca²⁺ uptake correlating with higher mtDNA.

We used a photoperiod manipulation of captive gray catbirds to investigate the role of SLN in the development of the migratory phenotype. We hypothesized that SLN mediates SERCA uncoupling in migratory songbirds that may lead to increased cytosolic Ca²⁺ transients in the muscle and signal for upregulated PGC1α (Maurya et al., 2018; Sopariwala et al., 2015), leading to muscle fatty acid catabolism, mitochondrial biogenesis and muscle hypertrophy (Fajardo et al., 2017; Summermatter et al., 2011). In our study, all birds demonstrated migratory restlessness after exposure to a LD photoperiod, with significant fat stores accumulating within 2 weeks. As expected, SLN was strongly upregulated in the migratory condition while SERCA2 remained stable. However, SLN transcription did not correlate with SERCA activity and Ca²⁺ dynamics, and while SLN showed a strong positive relationship with total fat stores, h-FABP and PGC1α across photoperiod treatments, there was no detectable relationship between SLN and pectoralis mass or mitochondrial biogenesis. Therefore, SLN appears to be strongly related to fat stores without the expected effects on SERCA activity and Ca²⁺ uptake, providing evidence that SLN function in songbirds deviates from what is seen in mammals.

We hypothesized that increasing the SLN/SERCA2 ratio in the muscle of birds in the migratory condition facilitates a shift to the endurance migratory phenotype through altered Ca²⁺ levels and subsequent PGC1α signaling (Maurya et al., 2018; Sopariwala et al., 2015). While the effects of SLN on SERCA activity and efficiency have been thoroughly examined in mammals, few studies have explored SERCA in birds (Pani et al., 2023; Elowe and Gerson, 2022; Pani and Bal, 2022; Price et al., 2019; Stager and Cheviron, 2020). To our knowledge, this study is the first to measure both SERCA activity and Ca²⁺ uptake in bird muscle, allowing us to detect uncoupling of Ca²⁺ pumping from ATP hydrolysis. We found that SLN transcription was significantly higher in the migratory condition while SERCA2 transcription remained stable. However, we found no relationship between SLN and SERCA activity or Ca²⁺ uptake, both of which respond to SLN uncoupling in mammals (Fajardo et al., 2013; Tupling et al., 2002). We also found no change in membrane permeability to Ca²⁺ (measured as the ratio of SERCA activity with versus without the Ca²⁺ ionophore) or SERCA efficiency (the apparent coupling ratio) with SLN transcription. This suggests that, if SLN transcripts translate to changes in protein abundance, SLN may not be a substantial uncoupler of SERCA. In support of this, Stager and Cheviron (2020) found that dark-eyed juncos (Junco hyemalis) were more susceptible to hypothermia when they had higher pectoralis muscle SLN mRNA, indicating that SLN likely did not contribute to SERCA inefficiency and, subsequently, non-shivering thermogenesis (NST) in the cold. Therefore, SLN may not appreciably affect thermogenesis or energy expenditure through SERCA inefficiency in birds.

While SLN did not appear to alter Ca²⁺ handling in the muscle, we did find a strong correlation between SLN, h-FABP and PGC1α mRNA, suggesting that SLN is associated with the shift to fat as a fuel. Moreover, while SLN overexpression is associated with reduced obesity in rodent models (Maurya et al., 2015), we found that SLN mRNA positively correlated with fat mass, providing further evidence that SLN may not function the same way in birds as in mammals. Given that these patterns were observed among individuals from both photoperiod treatments, transcription of SLN could be either a cause or a consequence of fat storage rather than indicative of the transition to a migratory phenotype, specifically. For example, greater amounts of stored fat could lead to adipokine signaling (Stuber et al., 2013), resulting in mitochondrial biogenesis and a shift to fat oxidation mediated through PGC1α (Babaei and Hoseini, 2022; Fang et al., 2023), which could lead to changes in Ca²⁺ signaling and SLN expression (Summermatter et al., 2011). Furthermore, as PGC1α signaling can be altered by a wide range of stimuli, including energy-sensing AMPK and stress-induced p38 MAPK (Miller et al., 2019), a number of other factors could alter PGC1α and, subsequently, SLN expression. Therefore, given the lack of discernible changes to Ca²⁺ dynamics in the muscle, the alterations to SLN expression may instead be a result of PGC1α changes induced by other stimuli, such as fatty acid signaling from increasing lipid stores (Supruniuk et al., 2017).

Given the strong relationship between SLN transcription and fat stores, it is surprising that there is no correspondence with fatty acid transporter CD36 or β-oxidation enzyme HOAD expression. Previous studies have shown that CD36 mRNA levels did not change in captive white-throated sparrows in the migratory condition (Price et al., 2010; Elowe et al., 2023a,b) or in wild-caught gray catbirds during spring migration (DeMoranville et al., 2019), though wild-caught migratory sparrows did show higher CD36 transcription (McFarlan et al., 2009) and CD36 also responds to flight activity (McFarlan et al., 2012; DeMoranville et al., 2020). Additionally, Zhang et al. (2015b) demonstrated in migratory yellow-rumped warblers (Setophaga coronata) and warbling vireos (Vireo gilvus) that while CD36 transcription was higher, this did not translate to higher protein levels. This may be the case for HOAD transcription in our catbirds as well, given that several studies have shown an increase in HOAD enzyme activity in the migratory condition (Marsh, 1981; McFarlan et al., 2009; Zhang et al., 2015b). However, other studies have also failed to detect changes to HOAD activity in the migratory birds (DeMoranville et al., 2019; Price et al., 2010). Therefore, CD36 and HOAD may not be consistent indicators of unexercised migratory condition.

Another hypothesized outcome of SLN-mediated PGC1α signaling is flight muscle hypertrophy, but we found no relationship between SLN transcription and pectoralis mass. While photostimulated captive white-throated sparrows tend to increase flight muscle mass (Elowe and Gerson, 2022; Price et al., 2011) and wild catbirds show pectoralis hypertrophy during migration (DeMoranville et al., 2019; Marsh, 1984), flight muscle mass may respond in part to exercise (Price et al., 2011; Zhang et al., 2015a). As with our birds, DeMoranville et al. (2019) found that pectoralis mass in wild catbirds was comparable between spring and winter, with highest mass during autumn migration. Therefore, pectoralis mass increase may not be a consistent indicator of the migratory condition and does not appear to be an outcome of increased SLN transcription.

We also hypothesized that SLN uncoupling would lead to signaling for mitochondrial biogenesis in the muscle, but we found no relationship between SLN transcription and SERCA coupling or mitochondrial DNA copy number, an endpoint estimate of mitochondrial biogenesis. Therefore, it does not appear that SLN facilitated any previous mitochondrial biogenesis via SERCA uncoupling in our study. However, without considering SLN, we still found a trend toward SERCA uncoupling leading to mitochondrial biogenesis regardless of photoperiod treatment, and this was primarily driven by a significant relationship with the rate of Ca²⁺ uptake. This provides evidence that SERCA uncoupling may lead to Ca²⁺ signaling for mitochondrial biogenesis (Maurya et al., 2018), but SLN does not appear to be a driver of this in our study.

The lack of evidence for SLN-mediated effects on SERCA uncoupling raises questions about the function of SLN in birds. Given the strong positive correlation between SLN transcription and fat mass, one possibility is that SLN exerts fat-dependent effects; if SLN increases with fat stores but its uncoupling effects are not consistent with its transcription, an additional factor may mediate the effect of SLN on SERCA in response to existing fat stores. For example, in rabbits and pigs, Montigny et al. (2014) found that depalmitoylation of a C⁹ residue on SLN led to a 30% increase in Ca²⁺-ATPase activity. This F⁹→C⁹ substitution is only seen in some mammals, but is consistent in passerine bird SLN sequences (Table S1). Migratory birds tend to increase levels of circulating fatty acids, including palmitate, during migration (Gupta et al., 2020), and this substrate is reversibly attached to cysteine residues on proteins either enzymatically or non-enzymatically (Qu et al., 2021). This could provide a nutrient-responsive mechanism to modulate SLN's effects as fat stores change. Furthermore, modulation of SLN's effects on SERCA may help to explain the inconsistency with mammalian models where SLN expression leads to reduced obesity (Maurya et al., 2015). However, without evidence that SLN affects SERCA in the expected ways in these birds, we may consider alternative reasons for SLN upregulation, such as SERCA stabilization in response to excess heat generation (Fu et al., 2020) during extended flapping flight.

Additionally, the timing and extent of SLN protein abundance changes may be important to consider. While SLN transcription was higher in the LD birds, mRNA is subject to post-transcriptional control (Mata et al., 2005) and degradation pathways (Boisvert et al., 2012) that could lead to substantial differences between SLN transcript and protein abundance and deviation between SLN transcription and SERCA uncoupling effects. Furthermore, we (Elowe and Gerson, 2022) found that SLN transcription in the flight muscle of white-throated sparrows was substantially higher when measured at 15 days rather than 25 days after LD photostimulation (nearly 50-fold over SD levels versus 20-fold, respectively), even though none of the birds exhibited Zugunruhe prior to day 15. Given that distinct transcriptomic patterns have been documented in birds between the different migratory fattening stages (Frias-Soler et al., 2022), SLN expression may be highest immediately after photostimulation to facilitate early physiological changes and reduced near-departure condition. If so, our measurements at 21 days after the photoperiod switch may be too late to document functional effects of SLN transcription. Furthermore, SLN may be responsive to the fattening associated with photoperiod stimulus, but its effects and expression as migration progresses from fattening to active defatting may respond to other cues, such as exercise. Recently, Valachovic et al. (2023) demonstrated that gray catbirds in a captive photoperiod stimulation of the autumn migratory condition gained large amounts of fat alongside reduced lipolysis in the adipose tissue, supporting the idea that fat breakdown constitutes a separate stage from photoperiod-induced premigratory fattening. Testing this would require a more in-depth sampling time course immediately following photostimulation and an evaluation of SLN protein abundance in the muscle tissue.

This study confirmed that photostimulated migratory birds show higher transcription of SLN in the absence of a corresponding change in SERCA2 transcription. This increase in SLN occurs during the transition to a migratory phenotype, including nocturnal migratory restlessness, fat accumulation and higher transcription of h-FABP, and may relate to the coordination of physiological changes by PGC1α. While this indicates that SLN may be involved in the process of storing and/or catabolizing fat stores, we found no evidence for SLN mediation of Ca²⁺ signaling through SERCA uncoupling and downstream effects on muscle hypertrophy and mitochondrial biogenesis. Therefore, further study is required to determine the extent to which SLN transcription leads to changes in protein abundance and interaction with SERCA isoforms, the timing of SLN expression in relation to photoperiod stimulation and migratory fattening, and how SLN may relate to the process of fat storage in birds.

We are grateful to all of the members of the UMass Avian Ecological Physiology Lab and Stager Lab who helped with this project. Notably, we thank Madelyn Kaplin, Anna De La Cruz, James Tsalah, Caitlyn Glidden, Jillian Zahar, Haley Bernardo and Abigail Prisby for help with animal care, along with University of Massachusetts Amherst Animal Care Services for their assistance. Thanks also to Julianna Applegate and Q. Salako for their help in conducting assays in the lab. We are grateful for the assistance from Russell Tupling and Val Fajardo in troubleshooting the SERCA assay and to Stephen McCormick, Courtney Babbitt and Matthew Fuxjager for suggestions on the manuscript.