Lung branching morphogenesis is accompanied by temporal metabolic changes towards a glycolytic preference

Background Lung branching morphogenesis is characterized by epithelial-mesenchymal interactions that ultimately define the airway conducting system. Throughout this process, energy and structural macromolecules are necessary to sustain the high proliferative rates. The extensive knowledge of the molecular mechanisms underlying pulmonary development contrasts with the lack of data regarding the embryonic lung metabolic requirements. Here, we studied the metabolic profile associated with the early stages of chicken pulmonary branching. Methods In this study, we used an ex vivo lung explant culture system and analyzed the consumption/production of extracellular metabolic intermediates associated with glucose catabolism (alanine, lactate, and acetate) by 1H-NMR spectroscopy in the culture medium. Then, we characterized the transcript levels of metabolite membrane transporters (glut1, glut3, glut8, mct1, mct3, mct4, and mct8) and glycolytic enzymes (hk1, hk2, pfk1, ldha, ldhb, pdha, and pdhb) by qPCR. ldha and ldhb mRNA spatial localization was determined by in situ hybridization. Proliferation was analyzed by directly assessing DNA synthesis using an EdU-based assay. Additionally, we performed western blot to analyze LDHA and LDHT protein levels. Finally, we used a Clark-Type Electrode to assess the lung explant's respiratory capacity. Results Glucose consumption decreases, whereas alanine, lactate, and acetate production progressively increase as branching morphogenesis proceeds. mRNA analysis revealed variations in the expression levels of key enzymes and transporters from the glycolytic pathway. ldha and ldhb displayed a compartment-specific expression pattern that resembles proximal–distal markers. In addition, high proliferation levels were detected at active branching sites. LDH protein expression levels suggest that LDHB may account for the progressive rise in lactate. Concurrently, there is a stable oxygen consumption rate throughout branching morphogenesis. Conclusions This report describes the temporal metabolic changes that accompany the early stages of chicken lung branching morphogenesis. Overall, the embryonic chicken lung seems to shift to a glycolytic lactate-based metabolism as pulmonary branching occurs. Moreover, this metabolic rewiring might play a crucial role during lung development. Supplementary Information The online version contains supplementary material available at 10.1186/s13578-021-00654-w.

lung development and defines the respiratory airway structure [1]. In the chicken, Gallus gallus, the primordial lung appears around day 3 of embryogenesis as a protuberance from the primitive foregut [2]. During this process, the mesobronchus grows distally, and the new secondary bronchi sprout laterally into the surrounding mesenchymal compartment [1,3]. This lateral or monopodial branching is exceptionally similar to the domain branching subroutine characteristic of the mammalian lung system [4,5]. Moreover, the molecular events underlying the development of the avian respiratory system are highly conserved with the mammalian and point to similar functions [6]. For instance, FGF (Fibroblast Growth Factor), WNT (Wingless-related Integration Site), SHH (Sonic Hedgehog), and Retinoic Acid signaling pathways were described as playing critical roles in chicken pulmonary branching morphogenesis [7][8][9][10][11].
The signaling mechanisms involved in early lung development are quite well studied in several animal models [10,12,13]; however, little is known concerning the embryonic lung metabolic needs [14,15]. As for the mammalian adult lung metabolic requirements, it has been shown that they are achieved through the uptake and catabolism of glucose, which represents the primary fuel to the adult lung tissue [16][17][18]. The coordination between signaling and metabolism is now emerging as a key concept for understanding developmental processes. Therefore, it is of major importance to investigate how metabolism contributes and is dynamically regulated during animal development [19,20].
Recent studies have demonstrated that glycolysis can serve additional roles beyond the classical bioenergetics purpose and contribute to shaping embryonic development both in time and space. For instance, during the murine chorioallantoic branching stage, the developing embryo redirects glucose carbon flow into the pentose phosphate pathway by suppressing phosphofructokinase 1 (PFK1) and aldolase; concomitantly, there is an increase in the glycolytic fraction to serve lactate biosynthesis. This study suggests a rewiring of glycolytic metabolism in the whole embryo, and over time, to promote accurate chorioallantoic branching [21]. Moreover, Slaninova et al. showed that loss of NOTCH signaling in Drosophila and human microvascular cells resulted in the downregulation of pivotal glycolytic genes; in opposition, a short pulse of NOTCH signaling stimulates glycolysis and increases lactate production, in detriment of Tricarboxylic Acid Cycle (TCA) activity [22]. Overall, these results revealed that NOTCH signaling promotes a glycolytic shift that resembles the Warburg effect [22]. Furthermore, Drosophila larval neuroblast cells, while proliferating, depend on aerobic glycolysis. However, during cell cycle exit and terminal differentiation, cell metabolism switches from aerobic glycolysis to an OXPHOS-based metabolism [23]. More recently, Bulusu et al. described a glycolytic activity gradient that contributes to Presomitic Mesoderm (PSM) differentiation and development in the mouse embryo [24]. Additionally, it has been shown that there is an FGF/Wnt coordinated glycolytic gradient that regulates cell motility and controls specification, thus contributing to PSM development in the chicken embryo [25].
This study aimed to investigate the metabolic profile underpinning the early stages of pulmonary branching morphogenesis. Chicken embryonic lungs were used to perform ex vivo lung explant culture. Extracellular metabolites associated with glucose catabolism (alanine, lactate, and acetate) were evaluated by 1 H Nuclear Magnetic Resonance ( 1 H-NMR) spectroscopy in the culture medium. The expression patterns/levels of key enzymes and transporters from the glycolytic pathway were assessed by in situ hybridization and qPCR, namely: glucose transporters (glut1, glut3, and glut8), monocarboxylate transporters (mct1, mct3, mct4, and mct8), and metabolic-related enzymes such as hexokinase (hk1 and hk2), phosphofructokinase (pfk1), lactate dehydrogenase (ldha and ldhb) and pyruvate dehydrogenase (pdha and pdhb). Proliferation status was determined by directly assessing DNA synthesis using an EdU-based assay. Additionally, the protein expression levels of LDH were determined by Western blot. Lastly, the respiratory capacity of explants was evaluated by measuring the basal oxygen consumption rate (OCR).
This report describes, for the first time, the temporal metabolic changes that accompany the early stages of chicken lung branching morphogenesis. In our experimental setting, the embryonic chicken lung shows a glycolytic preference with a shift to lactate production as pulmonary branching proceeds.

Lung branching morphogenesis is accompanied by temporal metabolite changes
To describe the metabolic alterations that occur during the early stages of chicken pulmonary branching, in vitro lung explant culture was performed using stages b1, b2, and b3 (1, 2, or 3 secondary buds formed per bronchus, respectively) that correspond to the first three branching stages (Fig. 1a); new branches are clearly seen with the epithelial marker l-cam (Fig. 1a). Explant culture allows assessing extracellular metabolite fluctuations while in a controlled environment. Briefly, the culture system was performed for 48 h and refreshed at D1 (24 h); medium was collected at D0 (0 h; to be used as reference/control), D1, and D2 (48 h), and then analyzed by 1 H-NMR spectroscopy. Metabolite production/consumption was a Representative examples of embryonic chicken lungs at DO (0 h) (b1, b2, b3) and D2 (48 h) (B1, B2, B3) of explant culture, and corresponding in situ hybridization for l-cam, (n = 5/stage). l-cam is an epithelial marker, showing active branching formation. Scale bar: 500 µm. b Schematic representation of the glucose catabolism pathway and the potential pyruvate destinations. Blue labeling indicates the metabolites that were detected and quantified in the 1 H-NMR spectroscopy analysis. c Glucose consumption, d Alanine production, e Lactate production, and f Acetate production, during 48 h of explant culture. The medium was refreshed at D1 (24 h). Medium samples were collected at D1 (24 h) and D2 (48 h). D0 media samples were used as control. Metabolite consumption or production was calculated following the mathematical formula | (D1-D0) + (D2-D0) |. Metabolite data was normalized to the total amount of protein. Results are expressed as | mean |± SEM (n ≥ 5/stage). One-Way ANOVA and Fisher's LSD test were performed. Significantly different results are indicated as: *p < 0.05; **p < 0.01 calculated following the mathematical formula | (D1-D0) + (D2-D0) |, expressed in pmol, and normalized to the total amount of protein. From the 1 H-NMR spectra analysis, it was possible to detect the following metabolites: glucose, alanine, lactate, and acetate (Fig. 1b). Additionally, morphometric analysis was performed to assess branching, revealing a gradual increment of the epithelial compartment when comparing between stages (B1 vs B2 vs B3), which implies an increase in lung branching morphogenesis (Additional file 1: Figure S1).
hk1 transcript (Fig. 3a) is present in the three pulmonary stages studied and displays statistically significant differences between b2 and B2 (p < 0.05). Also, hk1 expression decreases from B2 to B3 (p < 0.05). hk2 mRNA levels are practically null in the embryonic chick lung (data not shown).
pfk1 expression levels are maintained between groups (Fig. 3b), except for a statistically significant increase between B2 and B3 (p < 0.05).
Concerning ldha transcript (Fig. 3c), an increase in the expression levels from b1 to B1 (p < 0.001) and from b3 to B3 is detected (p < 0.05). After 48 h, both B2 and B3 lungs display a decrease in ldha levels compared to B1 lungs (p < 0.05). ldhb (Fig. 3d) is present in all the stages and time points analyzed, and statistically significant differences were observed only between b1 and B1 lungs (p < 0.05).
For all transcripts, fold variation was expressed as the change that occurs from 0 to 48 h of explant culture (Figs. 2g and 3g). Major variations were observed in stages b1 and b2 for glucose transporters (glut1 and glut3) and monocarboxylate transporters (mct1 and mct8) (Fig. 2g). Still, the b3/B3 stages exhibited only minor variations in the same groups of transcripts (Fig. 2g). hk1 and ldha showed a potential synergism with symmetric behavior regarding the fold variation (Fig. 3g). ldhb revealed a fold variation decrease for all stages, from 0 to 48 h. Still, no major expression variations were observed in pfk1, pdha, and pdhb (Fig. 3g).

ldha and lhdb exhibit region-specific expression patterns in the embryonic lung
To study the spatial distribution of lactate dehydrogenase, the expression pattern of ldha and ldhb was characterized by in situ hybridization. Afterward, representative lungs from the three branching stages were processed for histological sectioning.
ldha mRNA is present in the proximal epithelium of the lung, namely in the trachea region (Fig. 4a, dark arrowhead; 4b, black rectangle). ldha transcript is also expressed in the more distal region of the main bronchus (Fig. 4a, dagger). However, it is not expressed in the main bronchus epithelium and secondary bronchi (Fig. 4c, black arrow and asterisk, respectively). This expression pattern is conserved among the three branching stages Significantly different results are indicated as: *p < 0.05; **p < 0.01; ***p < 0.001 but decreases throughout branching morphogenesis, namely for stage b3. Slide sectioning of the hybridized lungs confirmed the presence of ldha in both mesenchymal and epithelial compartments of the embryonic trachea (Fig. 4d).
ldhb transcript is completely absent from the proximal region of the lung (Fig. 4e, dark arrowhead) and the epithelium of the primary bronchus (Fig. 4f, black arrow). ldhb is strongly expressed in the distal-most region of the lung (Fig. 4e, dagger) and the secondary bronchi ( Fig. 4g, asterisk). This expression pattern is maintained in the three stages studied. Lung sectioning confirmed that ldhb is not expressed in the distal epithelial compartment but exclusively present in the distal mesenchyme of the growing tips (Fig. 4h, dagger).

Lung active branching sites are associated with high proliferation
To assess the proliferation status of lung branching morphogenesis, we performed an EdU-based proliferation assay using B1 to B3 lung explants (Fig. 1a). Proliferation was determined by directly assessing EdU incorporation into new DNA strands using Alexa Fluor 488 (Green); nuclei were counterstained with Hoechst 33342 (Red) (Fig. 5).
After 48 h of culture, high proliferation levels were detected in the trachea region (white arrowhead) (Fig. 5b), in the distal tip of the lung (white arrow) (Fig. 5e), and in the secondary bronchi/active branching sites (asterisk) (Fig. 5h). The proliferation patterns are maintained in the three stages (B1, B2, and B3).

LDH protein levels support the increase of lactate production throughout branching
To understand how LDH protein levels may contribute to the observed increase in lactate production, the protein expression levels of LDHA and total LDH (that recognizes both LDHA and LDHB proteins) were assessed.
As seen in Fig. 6a, at 0 h, LDHA protein levels fluctuate between stage/condition. Semi-quantitative analysis ( Fig. 6b) revealed that LDHA expression levels increase at 0 h, namely from b1 to b3 (p < 0.01). On the other hand, after 48 h of culture, this variation dissipates, and there are no differences between stages. Conversely, when comparing both time points for each stage individually, a statistically significant decrease in LDHA expression levels is observed (Fig. 6b), namely from b1 to B1 (p < 0.01), from b2 to B2 (p < 0.0001), and from b3 to B3 (p < 0.0001).
In contrast, total LDH (LDHA plus LDHB) displays similar levels between stages at 0 h (Fig. 6c). Semi-quantitative analysis (Fig. 6d) exposed that there are no statistically significant differences between the three stages analyzed at 0 h. After 48 h of culture, total LDH protein levels vary between stages, namely from B1 to B2 (p < 0.05) and from B1 to B3 (p < 0.01). When comparing both time points for each stage individually, there is a statistically significant increase in total LDH expression levels from b2 to B2 (p < 0.001) and from b3 to B3 (p < 0.0001); in contrast, from b1 to B1 the protein levels remained unaltered.

Basal oxygen consumption rate is maintained during branching morphogenesis
To corroborate that the embryonic lung displays a glycolytic preference as branching morphogenesis develops, the respiratory capacity of explant tissue was evaluated. For this purpose, B1 to B3 lung explant tissue was collected after 48 h of culture (Fig. 1a), and the basal oxygen consumption rate was measured (Fig. 7).
After 48 h of culture, basal OCR was maintained between B1, B2, and B3 (Fig. 7). However, a slight decrease tendency in explants' respiratory capacity throughout pulmonary branching is detected.

Discussion
Lung development is a complex process characterized by epithelial-mesenchymal interactions mediated by signaling cascades, leading to the formation of a fully functional organ [12,26]. Altogether, these processes contribute to the formation of a fully functional adult organ responsible for gas exchange. The metabolic profile of the mammalian adult pulmonary system was studied in the '70 and '80 s, revealing a metabolism that is primarily dependent on glucose utilization [16][17][18]. On the other hand, little is known about embryonic lung metabolic needs. The importance of metabolism during development, beyond its canonical role, has been recently reported; several studies have shown that metabolism is highly regulated both in time and space, affecting multiple aspects of animal development [19]. Moreover, glucose uptake and catabolism are essential for all cellular processes and are believed to be enhanced in high proliferative systems such as cancer and embryonic development [20,27].
Branching morphogenesis is a crucial step in lung organogenesis since it establishes the future airway conducting system [28]. In the chicken lung, b1, b2, and b3 stages correspond to the sequential appearance of the first three secondary buds, with approximately an 8 h gap between them. These morphologically similar stages  Results are expressed as mean ± SEM (n ≥ 5/stage/condition). One-Way ANOVA and Fisher's LSD test were performed. p < 0.05 was considered revealed to be a useful system to study the molecular mechanisms underlying avian branching morphogenesis. Overall, signaling pathways display similar patterns between stages, and the interaction between different signaling events ultimately shapes airway branching [7][8][9]11]. In this sense, we wondered if b1, b2, and b3 stages were also similar at the metabolic level or if, despite the molecular similarities, the embryonic lung gradually adapts its metabolic needs to cope with the progressive tissue growth. Accordingly, in this work, we focused on describing the metabolic profile of the embryonic chicken lung during the abovementioned stages of branching morphogenesis. For this purpose, an ex vivo explant culture system was used to precisely measure the metabolic and associated molecular alterations during 48 h of culture [29][30][31]. This method is particularly relevant to study the whole organ while under very controlled conditions since it preserves the epithelialmesenchymal interactions that contribute to branching morphogenesis [30,32]. Lung explants are described to maintain the native physiological interactions between cells and tissues, thus mimicking the in vivo structure and function [29]. The culture medium from ex vivo lung explant culture was collected to study the production/consumption of extracellular metabolites related to glucose catabolism (alanine, lactate, and acetate) by 1 H-NMR spectroscopy. Afterward, the molecular machinery associated with glucose catabolism (glut1, glut3, glut8,  mct1, mct3, mct4, mct8, hk1, hk2, pfk1, ldha, ldhb, pdha, and pdhb) was characterized by qPCR. ldha and ldhb mRNA spatial localization was assessed by in situ hybridization. Proliferation status was determined by directly assessing DNA synthesis using an EdU-based assay. LDH protein expression levels were evaluated by western blot. The respiratory capacity of lung explants was evaluated by measuring the basal oxygen consumption rate.
In this study, 1 H-NMR results revealed that glucose consumption by chicken lung explants decreases from B1 to B2 and from B1 to B3 stage, and remains stable between B2 and B3 stage (Fig. 1c). In B1, glucose consumption is higher, and we believe this is due to a higher nutrient requirement to cope with the very beginning of branching morphogenesis. As lung branching morphogenesis proceeds, glucose consumption profile variation may reflect the need to cope with specific energy and nutrients requirements needed to maintain high proliferative rates. Similarly, the adult mammalian lung metabolic requirements are mainly achieved through glucose catabolism, although other metabolic substrates might support lung energetic demands [16][17][18]33].
Glucose cellular uptake relies on the presence of integral membrane glucose transporters (GLUTs) that move glucose from the extracellular space to the cytosol by facilitated diffusion; furthermore, this process depends on the number of glucose transporters expressed in the cell surface. In this sense, the expression levels of selected transporters, glut1, glut3, and glut8, in the chicken developing lung were characterized. glut1, glut3, and glut8 are expressed in the three pulmonary stages studied (Fig. 2b-d). After 48 h in culture, major variations of glut1 and glut3 are detected in B1 and B2 stages (Fig. 2b,  c) but not in the B3 stage (Fig. 2b, c). glut1 and glut3 expression level variations may support the observed glucose consumption variations. glut1 and glut3 transcripts have already been identified during chicken embryonic development in other tissues than the lung, namely, in the brain, muscle, and heart; moreover, glut1 and glut3 expression levels are developmentally regulated in these structures [34]. In the adult, glut1 and glut8 are ubiquitously expressed, whereas glut3 is highly expressed in the brain, similarly to their mammalian counterparts [35][36][37]. In human fetal lung, GLUT1 is present in bronchial and primitive alveolar epithelial cells during the branching phase but becomes progressively less expressed from week 19 onwards [38]. GLUT1 is widely expressed in adult human tissues, and it is commonly overexpressed in several tumor types [39]. On its turn, GLUT3 is highly expressed in lung cancer and actively contributes to cellular glucose uptake and consequent high proliferation of pulmonary tumor cells [40]. Altogether glut1, glut3, and glut8 may likely contribute to sustaining glucose consumption throughout early pulmonary branching morphogenesis stages.
Once inside the cell, glucose is converted into glucose-6-Pi in a reaction catalyzed by hexokinase [41]. hk1 is expressed in the embryonic chicken lung during branching morphogenesis (Fig. 3a), whereas hk2 transcript is barely detected (data not shown).
PFK1 catalyzes a key step of glycolysis, the unidirectional conversion of fructose-6-Pi and ATP into fructose-1,6-Pi and ADP [41]. In the embryonic chicken lung, pfk1 transcript is present in all stages studied without major variations between groups (Fig. 3b). pfk1 expression levels may contribute to control the relative glucose flux into the glycolytic pathway throughout early branching.
From the glycolytic pathway, each molecule of glucose is converted into two molecules of pyruvate. In our explant culture medium, several pyruvate-derived metabolites (alanine, lactate, and acetate from acetyl-CoA) were detected (Fig. 1d-f ). During branching morphogenesis, the embryonic chicken lung progressively increases alanine production that is finally exported to the extracellular medium (Fig. 1d). Previous studies revealed that, in the adult mammalian lung, part of the glucose carbons serves for alanine production through pyruvate transamination [17,18,33]. Similarly, in prostate cancer, the production of significantly high levels of alanine supports the need for proliferating cells for protein synthesis and membranogenesis [42]. Moreover, in pancreatic ductal adenocarcinoma, alanine produced from the surrounding environment is used by tumor cells for biosynthetic purposes [43]. Thus, alanine is an important end product of glucose catabolism and, in the developing lung, may contribute to protein biosynthesis to support active pulmonary growth.
Lactate was detected at high concentrations in the explant culture medium. Indeed, in our culture conditions, there is a sharp increase in lactate production from stage B1 to B3, and from stage B2 to B3 (Fig. 1e), concurrently with the increase in branching (Additional file 1: Figure S1). The lactate/glucose ratio revealed that in B1 and B2 explants, 41% and 58% of glucose is directed to lactate production, respectively; in B3 explants, around 76% of glucose is converted into lactate. Likewise, in the mammalian adult lung, around 50% of the glucose carbons are metabolized into lactate [16][17][18]33]. Therefore, lactate production by the developing lung may be facilitating the uptake and incorporation of nutrients by promoting the activity of biosynthetic pathways to, for instance, form new biomass and regenerate NAD + required for maximal glycolytic flux [44]. The increase in alanine and lactate production is characteristic of high proliferative systems in which fast energy and macromolecules are necessary to sustain the growth.
Lactate dehydrogenase (LDH) is the enzyme responsible for the interconversion of pyruvate into lactate and NADH into NAD + . This enzyme is a tetramer of two types of subunits, LDHA (formerly LDH-M) and LDHB (formerly LDH-H), encoded by ldha and ldhb, respectively [45]. In this work, we characterized the expression levels of both ldha and ldhb in the embryonic chicken lung. ldha expression increases from b1 to B1 and from b3 to B3 (Fig. 3c). Regarding 48 h' time point, the expression levels are higher at B1 and then decrease substantially to B2 and B3 (Fig. 3c). On the other hand, ldhb expression levels decrease from 0 to 48 h for b1 stage (Fig. 3d). Although the relative expression levels of ldhb are greater than ldha, higher variations accompanying branching were observed in ldha (Fig. 3g). Afterward, we performed in situ hybridization to assess ldha and ldhb spatial localization. ldha and ldhb displayed region-specific expression patterns in the chicken developing lung. ldha is highly expressed in the proximal region of the lung (Fig. 4a and d) and non-existent in the main bronchus epithelium and secondary bronchi (Fig. 4c). In opposition, ldhb transcript is absent from the proximal region of the lung (Fig. 4e) but expressed in the distal region. ldhb is specifically expressed in the growing tips and secondary buds ( Fig. 4e and g), suggesting an association with highly proliferative regions. Interestingly, in the chicken and mouse embryo, ldhb is found mostly in the posterior (distal-most) region of the embryo, the tailbud region [25]. ldha and ldhb specific patterns remind a spatial distribution typical of proximal-distal markers.
To determine the proliferation status of lung branching morphogenesis, we performed an EdU proliferation assay, using B1 to B3 lung explants (Fig. 1a). High levels of proliferation were detected in the trachea region (white arrowhead) (Fig. 5b), matching ldha mRNA expression pattern (Fig. 4d). In addition, high levels of proliferation were also present in the secondary bronchi/active branching sites (asterisk) (Fig. 5h), and in the distal tips of the developing lung (white arrow) (Fig. 5e), which coincides with ldhb transcripts spatial localization (Fig. 4eh). These results point towards an association between ldha and ldhb expression localization and proliferation events occurring during branching morphogenesis.
After the mRNA expression studies, we performed western blot for LDHA and total LDH (LDHA and LDHB contribution). LDHA displayed a decrease between 0 and 48 h (Fig. 6a, b). In contrast, total LDH expression levels increased from 0 to 48 h of ex vivo culture (Fig. 6c,  d). From these results, we speculate that LDHB may be contributing to the increase of total LDH protein levels. When we compare mRNA and protein expression data, an inverse pattern between the same ldh/LDH isoform is observed (Figs. 3c vs 6b; Figs. 3d vs 6d). It is generally assumed that mRNA expression levels directly correlate with protein levels. However, several studies have demonstrated that this not always occurs [46,47]. Distinct factors may account for these differences, such as regulatory mechanisms at transcriptional and/or translational level, or mRNA and protein stability, that can collectively impact total protein levels. In addition, given western blot results and mRNA spatial localization, it is possible that during branching, distal regions produce more lactate to cope with high proliferation rates. The increase in LDH total levels may account for the observed rise in extracellular lactate levels, which seem to be a characteristic of the developing lung.
Since lactate exchange between the cytosol and the extracellular space depends on monocarboxylate transporters (MCTs), we decided to study MCTs in the developing lung. mct1 and mct8 expression levels behave similarly throughout branching (Fig. 2e, f ) and recapitulate GLUTs mRNA expression behavior. mct1 displays a considerable increase in stage b1 after 48 h in culture (Fig. 2e), suggesting that it may support intracellular lactate export. Likewise, mct3 and mct4 were evaluated but displayed very low expression levels (data not shown). MCT1 has been described in both human and rat fetal lung; additionally, MCT inhibition leads to a decrease in rat lung branching [38]. More recently, Oginuma and colleagues found that mct1 is expressed, in a graded manner, in the tailbud and posterior PSM of the chicken embryo. Moreover, chicken explants cultured with MCTs inhibitor α-cyano-4-hydroxycinnamic acid (CNCn) displayed increased intracellular lactate levels [48]. On the other hand, cancer cells are equipped with MCTs, mainly MCT1 and MCT4, to exchange lactate and regulate pH homeostasis [49]. MCT4 is the predominant monocarboxylate transporter for lactate export in glycolytic cancer cells; conversely, MCT1 is a passive transporter that can operate in both directions and, in cancer systems, it is described to facilitate lactate export [50]. In fact, MCT1 lactate transport depends on the intracellular/ extracellular concentrations of lactate and protons [50]. Chicken mct8, also known as thyroid hormone transporter, was previously described in the developing brain, retina, spinal cord, kidney, and testis of the chicken [51,52]. Moreover, mct8 knockdown impacts embryonic chicken development, meaning that mct8 expression is crucial from the very early stages of organogenesis [53]. However, MCT8 does not transport lactate, but it may transport other monocarboxylates and amino acids [54]. In the embryonic chicken lung, mct1 can be the transporter responsible for cellular lactate exchange.
Pyruvate can be transformed into acetyl-CoA through an irreversible enzymatic reaction catalyzed by pyruvate dehydrogenase (PDH) and then incorporated in the TCA cycle or converted into acetate [41]. Both isoforms of pyruvate dehydrogenase, pdha and pdhb, are expressed in the embryonic chicken lung (Fig. 3e,  f ). In the embryonic lung explant culture medium, we did not find detectable levels of metabolites from the Krebs cycle in the 1 H-NMR spectra. On the other hand, acetate was identified, and its production is increased in later stages of early pulmonary branching (Fig. 1f ). Altogether, these data suggest that part of pyruvate may be diverted to acetate, through acetyl-CoA, as pulmonary branching proceeds. In mammals, glucose-derived pyruvate can generate acetate, and this phenomenon is more pronounced when under conditions of hyperactive glucose metabolism [55]. Moreover, in the adult mammalian lung, glucose carbons can incorporate acetate for lipogenesis purposes and serve for surfactant production [17,18,33,56]. Thus, in the embryonic lung, acetate may also point to a hyperactive glucose metabolism and can be produced to incorporate the newly synthesized cellular membranes, a common mechanism of high proliferative cells.
Lastly, to assess whether the embryonic lung truly shows a glycolytic preference during branching morphogenesis, mitochondrial respiration was evaluated. For this purpose, the respiratory capacity of explants was assessed by measuring the basal oxygen consumption rate of the embryonic lungs (Fig. 7). After 48 h of culture, explants display a steady basal OCR in the three stages; nonetheless, there is a slight tendency to decrease as branching morphogenesis proceeds (Fig. 7). This phenomenon occurs in parallel to the progressive increase in alanine, lactate, and acetate production, previously described (Fig. 1d-f ). Altogether, these results suggest that, under aerobic conditions and with functional mitochondria, the embryonic lung seems to shift to a glycolytic lactate-based metabolism throughout branching morphogenesis.

Conclusions
This study describes the temporal metabolic changes that occur during early chicken pulmonary branching. Throughout this period, we observed changes in the metabolite profile that occur concurrently with variations in the expression levels of pivotal enzymes and transporters from the glycolytic pathway and with a steady oxygen consumption rate (Fig. 8). It appears that pulmonary branching morphogenesis progressively adapts to a glycolytic lactate-based metabolic profile, suggesting a Warburg-like metabolism. This metabolic rewiring to aerobic glycolysis is also observed in other high proliferative developing systems and allows the production of energy and biomass during embryonic development [20-22, 24, 25, 57]. Moreover, this metabolic adaptation might exert additional signaling functions throughout pulmonary branching, and further studies are required to address this topic. We acknowledge that this study has a strong descriptive perspective, but we are addressing a subject that has not been exploited so far. This report highlights the importance of metabolic regulation during the early stages of lung development and lays the groundwork for future mechanistic studies on this topic.

Ethical statement
This work was performed at the early stages of chicken embryonic development and, therefore, does not require ethical approval following the European Parliament

Tissue collection
Fertilized chicken eggs, Gallus gallus, were incubated between 4.5 and 5.5 days (Embryonic day 4.5-5.5) in a 49% humidified atmosphere at 37 °C (Termaks KB400, Norway). Embryonic lungs were obtained by microdissection under a stereomicroscope (Olympus SZX16, Fig. 8 Schematic representation of the metabolic profile of early pulmonary branching morphogenesis. The metabolites detected in the extracellular medium (blue labeling) can either be consumed (glucose) or produced (alanine, lactate, and acetate) by the lung tissue. The symbols refer to increase (↑) or decrease (↓) in the metabolites, considering their particular fate. Temporal metabolite changes are accompanied by variations in the expression of glut1, glut3, glut8, hk1, pfk1, mct1, mct8, ldha, ldhb, pdha, and pdhb. Decreased LDHA and increased LDHB protein levels may contribute to the observed increase in the extracellular levels of lactate. OXPHOS is active with mitochondria displaying a constant rate of oxygen consumption. The symbols refer to: does not contribute to lactate transport (*); oxidative phosphorylation is active (✓) Japan) and classified in stage b1, b2, or b3 according to the number of secondary buds formed per bronchus; 1, 2, or 3, respectively [7]. Dissected lungs were processed for ex vivo lung explant culture or for in situ hybridization.

H-NMR spectroscopy
Samples of 200 µL of medium were collected from ex vivo lung explant culture, at D0, D1, and D2, and analyzed by 1 H-NMR spectroscopy (n ≥ 5/stage/condition) according to [58]. A Bruker Avance 600 MHz spectrometer with a 5 mm QXI probe and z-gradient (Bruker Biospin, Germany) was used, and spectra accessed at 25 °C. Solvent-suppressed 1 H-NMR spectra were acquired with 6 kHz spectral width, 14-s interpulse, 3-s water presaturation, 45-degree pulse angle, 3.5-s acquisition time, and a minimum of 128 scans. Sodium fumarate (singlet, at 6.50 ppm) was used as an internal reference at 10 mM. The following metabolites were detected and quantified: H1-α glucose (doublet, 5.22 ppm), alanine (doublet, 1.46 ppm), lactate (doublet, 1.33 ppm), and acetate (singlet, 1.9 ppm). To quantify the relative areas of 1 H-NMR resonances, the curve-fitting routine supplied with the NUTSproTM NMR spectral analysis program (Acorn NMR, USA) was used. D0 media samples were used as reference/control. The metabolite consumption/production during 48 h of explant culture was calculated following the mathematical formula | (D1-D0) + (D2-D0) |. Metabolite data were normalized to the total amount of protein.

Quantitative PCR
Total RNA was extracted from D0 and D2 lung explants using TripleXtractor directRNA kit (Grisp, Portugal). RNA integrity and purity were determined using the Experion RNA StdSens Analysis Kit (Bio-Rad, USA), and RNA concentration determined by Nanodrop 1000 Spectrophotometer (Thermo Fisher Scientific, USA). Total RNA was treated with DNase I (Thermo Fisher Scientific) and then reversely transcribed to obtain cDNA using GRS cDNA Synthesis kit (Grisp). Specific exonexon spanning primers were designed for the amplification of target and housekeeping transcripts (Additional file 1: Table S1). Primers were optimized for annealing temperature and PCR cycles, using NZY Taq 2 × Green Master Mix (NZYTech, Portugal) and, subsequently, for efficiency range. qPCR was performed in duplicate (n ≥ 5/stage/condition) using 1 µL of cDNA and SYBR method according to manufacturer's instructions, NZY qPCR Green Master Mix (2x) (NZYTech). mRNA expression levels were normalized for both 18s and actin-β housekeeping genes. Data on the gene expression levels were calculated following the mathematical model 2^(-ΔCt) [59].

Histological sections
Hybridized chicken lungs were dehydrated through increasing ethanol series and embedded in a 2-hydroxyethyl methacrylate solution (Heraeus Kulzer, Germany). 25 µm thick histological slides were produced by a rotatory microtome (Leica RM 2155, Germany). Lung histological sections were photographed using a camera (Olympus DP70) coupled to a microscope (Olympus BX61).

Proliferation assay and confocal microscopy
After 48 h of culture, half of the explant's media was replaced by fresh media containing EdU, at a final concentration of 150 µM. Explants were incubated with EdU for 90 min in the same culture conditions. After incubation, tissues were fixed in PBS solution containing 3.7% formaldehyde. Then, tissues were washed in PBS with 3% BSA, and permeabilized in PBS with 0.5% Triton X-100, for 90 min. Tissues were washed again and processed for Click-iT Plus EdU reaction, according to the manufacturer's instructions (Click-iT ™ Plus EdU Cell Proliferation Kit for Imaging, Invitrogen). Detection of the incorporated EdU was performed using Alexa Fluor 488. Nuclei were counterstained with Hoechst 33342 (1:2000). Images were acquired using an Olympus LPS Confocal FV3000 microscope (Olympus).

Basal oxygen consumption rate measurements
Pooled samples of D2 (2 lungs per pool/per stage) explants were processed to assess respiratory capacity (n ≥ 5/stage/condition). Lung tissue was washed in PBS and incubated, for 2 min, in respiration medium, at 37 °C. Respiration medium was composed by medium 199 (5.5 mM glucose) supplemented with 1% L-glutamine, 1% ITS (0.01 mg/mL recombinant human insulin, 0.0055 mg/mL human transferrin substantially iron-free, and 0.005 μg/mL sodium selenite) (BD Biosciences, USA) and 10 mM HEPES; pH was adjusted to 7.2-7.4, at 37 °C. To assess Oxygen Consumption Rate (OCR), a Clark-type electrode Oxytherm System (Hansatech, UK) was used [62]. Samples were transferred in 1 mL respiration medium to the pre-calibrated and thermostatized (with a water jacket, at 37ºC) electrode chamber. Oxygen recording protocol was performed for 15 min with an open-chamber plus 15 min with a closed-chamber; data acquisition refers to the last 5 min of closed-chamber mode. After each assay, tissue samples were retrieved, washed in PBS, and processed for protein extraction and quantification. Data was analyzed using Oxytrace Plus acquisition software (Hansatech) and normalized to the total amount of protein.

Statistical analysis
Statistical analysis was performed using GraphPad Prism 6 (GraphPad Software, USA). The normality of distribution was tested using the Kolmogorov-Smirnov test. One-Way ANOVA was performed and followed by Fisher's Least Significant Difference (LSD) post hoc test for multiple comparisons. All experimental data are presented as mean ± standard error of the mean (SEM) with a statistically significant level of 5% (p < 0.05) considered.