Gyrfalcons Falco rusticolus adjust CTNS expression to food abundance: a possible contribution to cysteine homeostasis
- 229 Downloads
Melanins form the basis of animal pigmentation. When the sulphurated form of melanin, termed pheomelanin, is synthesized, the sulfhydryl group of cysteine is incorporated to the pigment structure. This may constrain physiological performance because it consumes the most important intracellular antioxidant (i.e., glutathione, GSH), of which cysteine is a constitutive amino acid. However, this may also help avoid excess cysteine, which is toxic. Pheomelanin synthesis is regulated by several genes, some of them exerting this regulation by controlling the transport of cysteine in melanocytes. We investigated the possibility that these genes are epigenetically labile regarding protein intake and thus contribute to cysteine homeostasis. We found in the Icelandic population of gyrfalcon Falco rusticolus, a species that pigments its plumage with pheomelanin, that the expression of a gene regulating the export of cystine out of melanosomes (CTNS) in feather melanocytes of developing nestlings increases with food abundance in the breeding territories where they were reared. The expression of other genes regulating pheomelanin synthesis by different mechanisms of influence on cysteine availability (Slc7a11 and Slc45a2) or by other processes (MC1R and AGRP) was not affected by food abundance. As the gyrfalcon is a strict carnivore and variation in food abundance mainly reflects variation in protein intake, we suggest that epigenetic lability in CTNS has evolved in some species because of its potential benefits contributing to cysteine homeostasis. Potential applications of our results should now be investigated in the context of renal failure and other disorders associated with cystinosis caused by CTNS dysfunction.
KeywordsCTNS Cysteine homeostasis Food abundance Gyrfalcon Pheomelanin
Vertebrate pigmentation mainly depends on several genes that regulate the synthesis, structure and transport of melanins in melanocytes (Hubbard et al. 2010; Hearing 2011). During the synthesis of the sulphurated form of melanin (i.e., pheomelanin), which is responsible for orange-reddish animal coloration, the sulfhydryl group of the amino acid cysteine is incorporated to the pigment structure (García-Borrón and Olivares Sánchez 2011). As pheomelanin is then transferred to inert integumentary structures of the skin, feathers or hairs, cysteine used for pheomelanin synthesis is no longer available as an antioxidant as part of the tripeptide glutathione (GSH). As a consequence, pheomelanin synthesis represents a consumption of the most important intracellular antioxidant (Panzella et al. 2014). This consumption of cysteine/GSH contributes to some disorders, including melanoma and malignant glioma in humans (Panzella et al. 2014; Robert et al. 2015; Wang et al. 2016) and constraints to physiological performance in animals exposed to environmental sources of oxidative stress (Karell et al. 2011; Galván et al. 2011, 2014). However, cysteine is toxic if in excess; thus, its consumption during pheomelanogenesis may also be advantageous under environmental conditions leading to excess cysteine, which may indeed explain the evolution of pheomelanin-based pigmentation (Galván et al. 2012).
The expression of genes regulating pheomelanin synthesis should, therefore, be a key in determining the double-sided nature of pheomelanin-based pigmentation, in interaction with environmental factors. Some of us have recently found that zebra finches Taeniopygia guttata developing feathers pigmented by pheomelanin downregulate the expression of the gene Slc7a11 (solute carrier family 7 member 11), which codes for the plasma membrane cystine/glutamate exchanger xCT that transports cystine (the cysteine dimer) to melanocytes to produce pheomelanin (Chintala et al. 2005), under exposure to a pro-oxidant substance (Galván et al. 2017). As this downregulation of Slc7a11 avoided cellular oxidative stress exerted by the pro-oxidant, it means that some species have evolved an epigenetic lability in Slc7a11 that reduces cysteine consumption for pheomelanin production under stressful environmental conditions, when cysteine/GSH are most required to combat oxidative stress (Galván et al. 2017). Similar to this influence of environmental oxidative stress on gene expression, epigenetic permeability of genes regulating pheomelanin synthesis to environmental conditions that produce excess cysteine should also be physiologically advantageous. However, this possibility has never been explored under natural variations in dietary cysteine levels (i.e., food abundance) as a potential adaptive strategy.
Cysteine is a semi-essential amino acid; thus, diet is an important source of cysteine for animals (Klasing 1998; Stipanuk et al. 2006). Excess cysteine can occur when its content in diet is higher than needed for protein synthesis (Stipanuk et al. 2006). Under such conditions, the toxicity of excess cysteine is produced by its autoxidation to cystine, which generates hydrogen peroxide, hence decreasing GSH levels (Viña et al. 1983) and causing oxidative DNA damage (Park and Imlay 2003). This causes several diseases in mammals and metabolic acidosis and disorders such as thinning of egg shells and poor growth in birds (Klasing 1998). However, cysteine is essential for protein structure, hence the importance of maintaining cysteine homeostasis. The enzyme cysteine dioxygenase (CDO) catalyzes the addition of molecular oxygen to the sulfhydryl group of cysteine to form less toxic products such as sulfate and taurine (Stipanuk et al. 2006). Nevertheless, the capacity of CDO to maintain cysteine homeostasis is probably limited, as derived from the fact that cystinosis, a disease characterized by the intralysosomal accumulation of cystine, is only explained by dysfunction in the cystine/H+ symporter (cystinosin) that exports cystine out of lysosomes (Chiaverini et al. 2012). Pheomelanin synthesis may also affect, together with CDO, cysteine homeostasis (Galván et al. 2012). In fact, cystinosin and pheomelanin are intrinsically associated, as the gene encoding cystinosin (CTNS) is expressed in the organelles where melanin synthesis takes place in melanocytes (i.e., melanosomes) and, thus, inhibits pheomelanogenesis by promoting the efflux of cystine (Chiaverini et al. 2012).
Potential epigenetic mechanisms effective in avoiding excess cysteine may, therefore, be related to changes in the expression of Slc7a11 and CTNS. Here we test these possibilities in a wild population of gyrfalcons Falco rusticolus exposed to different food regimes in Iceland, as these birds synthesize pheomelanin to pigment the feathers of their first plumage (Galván and Jorge 2015). As the gyrfalcon is a strict carnivore, variation in the food abundance to which nestlings are exposed during growth mainly reflects variation in protein availability in the diet (Klasing 1998). We also considered the gene Slc45a2 (solute carrier family 45 member 2), which has also been suggested to affect pheomelanin synthesis by transporting cysteine into melanosomes (Xu et al. 2013). Lastly, we considered the melanocortin-1 receptor (MC1R) and agouti-related protein (AGRP) genes, which affect pheomelanin synthesis without influencing cysteine/GSH availability (Nadeau et al. 2008). We, thus, predicted that if epigenetic sensitivity to protein availability has evolved in Slc7a11 and CTNS (and possibly Slc45a2) as an adaptive mechanism to avoid excess cysteine, the expression of these genes should be adjusted to food abundance while MC1R and AGRP should not be affected.
The study was conducted during two consecutive breeding seasons (June 2015 and 2016) in 16 gyrfalcon territories in Iceland. Nests were entered when the nestlings were 16–35 days old. Nestlings were ringed, weighed and the following morphometrics taken: flat chord of the wing, culmen and length of central tail feathers. Sexing was based on morphometrics. Ageing was based on length of central tail feather (Ó.K. Nielsen, unpublished data). Body condition was calculated as the residuals of body mass regressed against wing chord, culmen length and age. The study comprised a total of 47 gyrfalcon nestlings (Online Resource 1).
Estimation of food abundance
Two or three collecting trips were made to each nest during the season, the last after the young had fledged. During each visit all remains were collected, including all bones and pellets, and also some feathers. For the analysis, remains were arranged according to species, age groups and type, from which we estimated the number and biomass of prey [see Nielsen (2003) for details]. Number and biomass of prey were thus considered as two indices of food abundance to which nestling gyrfalcons were exposed during growth.
Gene expression in melanocytes
During feather development, melanocytes at the dermal papillae show intense melanogenesis (Lin et al. 2013). Therefore, the bottommost portion of the feather follicles corresponding to the melanin unit, which represents an important reservoir of melanocytes like in hair follicles (Mohanty et al. 2011), was isolated. RNA obtained from these samples, therefore, corresponds to melanocytes to a large extent. All follicular melanin units obtained from each gyrfalcon nestling were pooled. Total RNA was extracted using TRI Reagent (Ambion), and residual genomic DNA carry over was removed using the TURBO DNA-free kit (Ambion). Complementary DNA (cDNA) was prepared from total RNA using RevertAid Reverse Transcriptase provided in the RevertAid First Strand cDNA Synthesis kit (Thermo Scientific, Thermo Fisher Scientific). qPCR was thus performed on cDNA for the genes CTNS, Slc7a11, Slc45a2, MC1R and AGRP. Reactions were performed using SYBR Green I Master in a LightCycler 480 System (Roche, Basel, Switzerland). The housekeeping glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was used for normalization of cDNA samples. Cycle threshold (C t) levels were used as a measure of gene expression, with lower C t values indicating higher mRNA levels and higher gene expression levels. Two replicates were obtained for each sample, and the mean value was used in the statistical analyses. We did not obtain any amplification of genes from eight nestlings in the case of CTNS, from 6 nestlings in the case of Slc7a11, from seven nestlings in the cases of Slc45a2, MC1R and GAPDH, and from ten nestlings in the case of AGRP (Online Resource 1).
Gene primers were designed based on Falco cherrug refseq sequences (GenBank) using the Primer-BLAST tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). The oligonucleotide primer sets used were as follows: 5′-GCTGGGGAATTGCTGCTTT-3′ and 5′-TGCAAC GAAGAACATCCTGG-3′ for Slc7a11; 5′-CATGCCATCCCACTATCGCT-3′ and 5′-GAGGCGCATAAGGACTACCC-3′ for Slc45a2; 5′-GGGAGAAGCCGTGAGCTTT-3′ and 5′-CAGTGAGTCCATCACCAGGC-3′ for CTNS; 5′-CTCCTTCCTGGGGGTCATC-3′ and 5′-TGATGAAGATGGTGCTGGAGA-3′ for MC1R; 5′-GAAGCCAACGGTGACCTCAT-3′ and 5′-AAGGCATTGAAGAAGCGGCA-3′ for AGRP; and 5′-GGACCAGGTTGTTTCCTGTG-3′ and 5′-TCCTTGGATGCCATGTGGA-3′ for GAPDH.
We tested if gene expression levels (C t) changed with food abundance using linear mixed-effect models (LMM) fit with restricted maximum likelihood (REML) estimation in R environment (R Core Team 2017). As the two indices of food abundance (i.e., number and biomass of prey) were strongly correlated (r = 0.98, N = 19, P < 0.0001), we conducted separate models for these variables. LMM’s thus included number or biomass of prey, in addition to sex and body condition, as fixed terms. Territory identity was added as a random factor to account for the common origin of gyrfalcon nestlings belonging to the same nests. Different models were fit for each gene. Effect sizes (b) were calculated with the library lme4, and P values were calculated from the analysis of deviance of the models using the library car. All continuous variables were log10-transformed prior to analyses to achieve normality assumptions.
Although modulation of CTNS expression in response to intracellular cysteine levels has been shown in vitro in human kidney cells (Bellomo et al. 2010), our findings represent the first evidence of epigenetic lability in the gene CTNS as response to a natural variation in a dietary source of cysteine. Diet constitutes an important source of cysteine, and consequently elicits the modulation of the enzyme CDO to maintain cysteine homeostasis (Stipanuk et al. 2009). Protein intake, therefore, affects overall cysteine levels in the organism (Stipanuk et al. 2009), especially in strict carnivores like the gyrfalcon. CTNS expression was highest in gyrfalcon nestlings reared in territories with highest food abundance, meaning that this gene is epigenetically labile in gyrfalcons. We thus suggest that epigenetic sensitivity to protein intake in CTNS may contribute to cysteine homeostasis, because a high expression of this gene avoids cystine accumulation in melanosomes (Chiaverini et al. 2012).
CTNS affects pheomelanin synthesis, whose induction may represent a mechanism to remove excess cysteine (Galván et al. 2012). However, CTNS only affects pheomelanogenesis under high systemic cysteine levels in the organism as derived from a comparison between levels of cystine in mouse blood (50 µM) and in B16 cell culture medium (380 µM), suggesting the existence of a threshold cysteine/cystine concentration below which cystinosin activity and pigmentation are unaffected (Chiaverini et al. 2012). While the threshold value will have to be investigated in birds, this may explain why CTNS expression increased and not decreased with food abundance in gyrfalcons. The possible physiological benefits derived from the epigenetic lability of CTNS may, therefore, be related to cysteine efflux instead of pheomelanin synthesis. This may, thus, contribute to cysteine homeostasis without affecting pigmentation. In fact, although we could not characterize the plumage pigmentation of nestling gyrfalcons, it is known that a single plumage color variant that is explained by certain MC1R alleles exists in the Icelandic population of this species (Johnson et al. 2012). In other words, Icelandic gyrfalcons show variation in the expression of CTNS in feather melanocytes according to food abundance in the breeding territories, but little variation in plumage pigmentation.
However, it is not expected that all species have evolved the capacity to adjust CTNS expression to dietary influences. In this regard, CTNS dysfunction causes cystinosis in humans and other animals, an autosomal recessive disease that has renal failure as the most severe symptom (Chiaverini et al. 2012). A specific gene therapy has proved successful at decreasing cystine levels in CTNS-deficient cells in mice (Harrison et al. 2013). This response would possibly be enhanced by an additional inducement of CTNS expression, which may be achieved by a mechanism similar to that allowing gyrfalcons to increase CTNS expression in response to food abundance. Although to our knowledge, no study has addressed the possibility that exposure to solar (UV) radiation affects the expression of pigmentation genes in birds, this is known to occur in humans regarding the MC1R gene (Scott et al. 2002); thus, the potential influence of UV exposure on CTNS should be considered together with diet. Future studies should investigate this potential applicability of CTNS epigenetic lability to avoid the physiological consequences of cystinosis in animals including humans.
It could also be suggested that the potential lack of dietary influence on CTNS expression in humans may represent a physiological cost for those producing large amounts of pheomelanin (i.e., associated with low CTNS expression; Chiaverini et al. 2012) such as humans with the red hair phenotype (Ito et al. 2011), thus contributing to some damaging consequences associated with this phenotype like a higher resistance to analgesics and an increased tendency to bruising (Liem et al. 2006). Some evidence suggests that Neanderthal hominids had a red hair phenotype (Lalueza-Fox et al. 2007), and it is known that their diet was heavily based on meat in at least some populations (Weyrich et al. 2017), thus allowing to speculate that a low CTNS expression with no epigenetic lability may have contributed to their extinction. Our study should stimulate research on the new field on the interaction between epigenetic lability and pigmentation phenotype in a variety of study areas.
Two anonymous reviewers made useful suggestions that improved the manuscript. Sara Borrego and Rafael Márquez helped with laboratory analyses and Ólafur H. Nielsen, María Nielsen, Einar Þorleifsson and Kristinn P. Magnússon with fieldwork. ÓKN has state permit to access gyrfalcon nests, ring chicks and to collect non-invasive genetic samples. IG is supported by a Ramón y Cajal Fellowship (RYC-2012-10237) and the Project CGL2015-67796-P, both from the Spanish Ministry of Economy and Competitiveness (MINECO).
Author contribution statement
IG conceived and designed this study, conducted laboratory work, analyzed the data and wrote the manuscript. ÂI designed gene primers and participated in laboratory work. ÓKN conducted fieldwork.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
Statement of animal rights
All applicable institutional and/or national guidelines for the care and use of animals were followed.
- Ito S, Nakanishi Y, Valenzuela RK, Brilliant MH, Kolbe L, Wakamatsu K (2011) Usefulness of alkaline hydrogen peroxide oxidation to analyze eumelanin and pheomelanin in various tissue samples: application to chemical analysis of human hair melanins. Pigment Cell Melanoma Res 24:605–613CrossRefPubMedGoogle Scholar
- Klasing KC (1998) Comparative avian nutrition. CAB International, WallingfordGoogle Scholar
- Nielsen ÓK (2003) The impact of food availability on gyrfalcon (Falco rusticolus) diet and timing of breeding. In: Thompson DBA, Redpath SM, Fielding AH, Marquiss M, Galbraith CA (eds) Birds of prey in a changing environment. Scottish Natural Heritage, Edinburgh, pp 283–302Google Scholar
- R Core Team (2017) R: a language and environment for statistical computing. R Foundation for Statistical Computing, ViennaGoogle Scholar
- Robert SM, Buckingham SC, Campbell SL, Robel S, Holt KT, Ogunrinu-Babarinde T, Warren PP, White DM, Reid MA, Eschbacher JM et al (2015) SLC7A11 expression is associated with seizures and predicts poor survival in patients with malignant glioma. Sci Transl Med 7:289ra86CrossRefPubMedPubMedCentralGoogle Scholar