ABSTRACT
Slc7a7 encodes for y+LAT1, a transporter of cationic amino acid across the basolateral membrane of epithelial cells. Mutations in SLC7A7 gene give rise to Lysinuric Protein Intolerance (LPI), a rare autosomal recessive disease with wide variability of complications. Intriguingly, y+LAT1 is also involved in arginine transport in non-polarized cells such as macrophages. Here we report that complete inducible Slc7a7 ablation in mouse compromises systemic arginine availability that alters proper erythropoiesis and that dysfunctional RBC generation leads to increased erythrophagocytosis, iron overload and an altered iron metabolism by macrophages. Herein, uncovering a novel mechanism that links amino acid metabolism to erythropoiesis and iron metabolism. Mechanistically, the iron exporter ferroportin-1 expression was compromised by increased plasma hepcidin causing macrophage iron accumulation. Strikingly, lysozyme M-cell-specific knockout mice failed to reproduce the total knockout alterations, while bone marrow transplantation experiments resulted in the resolution of macrophage iron overload but could not overcome erythropoietic defect. This study establishes a new crucial link between systemic arginine availability in erythropoiesis and iron homeostasis.
Introduction
Red blood cell (RBC) generation is a tightly regulated process where RBC homeostasis is key for proper iron recycling (de Back et al., 2014). Although extensive work has been done in the field of erythropoiesis, little is known about the impact of amino acid metabolism in this complex process. The mechanisms of RBC generation spans from bone marrow (BM) erythroid differentiation mediated by CD169+ macrophages (Chow et al., 2013), which supports erythroblastic island formation, to the end-final stage where RBC phagocytosis by red pulp macrophages (RPMs) leads to hemoglobin breakdown and ultimately iron recycling and release (Klei et al., 2017). RBC maturation requires specific components to properly coordinate this process. Disruption in hemoglobin synthesis, which comprises one-third of the RBC protein content, leads to altered erythropoiesis (Kuhn et al., 2017; Liu et al., 2013) .Yet, hemoglobin is not the only key component as iron, erythropoietin (EPO), or ferritin are also well known to play essential roles in RBC generation (Beguin, 1998; Goldfarb et al., 2021; Moritz et al., 1997). In terms of metabolic requirements, Shima et al. published the impact of arginine import on erythrocyte differentiation and proliferation throughout the cationic amino acid transporter 1 (CAT1), thereby indicating a crucial role not only of the iron-related components (hemoglobin, EPO, iron and ferritin) but also of metabolites such as arginine in the generation of mature RBCs (Shima et al., 2006). Macrophages also play an important role in RBC enucleation, being thus key for the last step of RBC generation (Lee et al., 2004; Popova et al., 2009; Swartz et al., 2017).
Macrophages are a cell type that participates in diverse biological processes, including host defence and wound repair (Koh and DiPietro, 2011). Nevertheless, further roles for these cells began to emerge with the identification of specific functions of tissue-resident macrophages, such as Kupffer cells and splenic RPMs, which are mainly involved in erythrocyte phagocytosis and iron recycling (Beaumont and Delaby, 2009; Ganz, 2012; Theurl et al., 2016), alveolar macrophages (AMs), which participate in both lung development and surfactant recycling, and osteoclasts, which contribute to bone development (Hussell and Bell, 2014; Murray and Wynn, 2011). Interest in macrophages in the context of metabolic disease has gained momentum due to a number of recent findings. Macrophage polarization is well known to be tighly linked to altered cellular metabolism including iron homeostasis and glycolysis/citric acid cycle activity (Recalcati et al., 2012; Stienstra et al., 2017). In addition, changes in L-arginine metabolism have been coupled to different immune effector phenotypes of macrophages involved in autoimmunity, infection control and activation (Bronte and Zanovello, 2005; Jha et al., 2015; Weiss and Schaible, 2015).
Lysinuric Protein Intolerance (LPI, MIM 222700) is a rare autosomal recessive disease caused by mutations in SLC7A7 gene (solute carrier family 7) which encodes for y+LAT1 (Palacín et al., 2001; Torrents et al., 1999) , a light subunit of the heterodimeric amino acid transporter family. Y+LAT1 mediates the exchange of cationic amino acids (CAAs) with neutral amino acids plus sodium (Palacín et al., 2005) across the basolateral membrane of epithelial cells. Mutations in y+LAT1 results in defective transport of CAAs, leading to reduced arginine, ornithine and lysine plasma concentration while increased in urine (Ogier de Baulny et al., 2012). Consistent with the clinical manifestations of human LPI, we have previously reported that the inducible complete loss of y+LAT1 in mice leads to hypoargininemia, which results in urea cycle disruption and hyperammonemia. Consequently leading to reduced body weight, brain edema and pulmonary alveolar proteinosis between other complications (Bodoy et al., 2019). In addition, several studies found that patients with one or several mutations in Slc7a7 gene have abnormal blood count, as well as microcytic anemia (Alqarajeh et al., 2020; Rajantie et al., 1980). To date, the standard treatment for LPI mainly consists on a low-protein based diet supplemented with oral citrulline (Lukkarinen et al., 2003), where citrulline is intracellularly converted to arginine in renal epithelial cells. Hence, improving the defects in urea cycle and correcting both plasma arginine and ammonia levels (Dhanakoti et al., 1990).
Notably, y+LAT1 also mediates arginine transport in non-polarized cells, such as macrophages. Intriguingly, y+LAT1 was shown to drive major arginine transport in human monocytes after interferon stimulation (Rotoli et al., 2020). Thus, being one of the major arginine transporters in human monocytes, AMs and monocyte-derived macrophages (Barilli et al., 2012).
Motivated by the fact that human SLC7A7 mutations give rise to immune and hematological complications, here we questioned whether amino acid transport via Slc7a7 has important roles for erythropoiesis and/or iron homeostasis. Using total loss-of-function of y+LAT1, recovery with citrulline, Lysozyme M-cell-specific (i.e. myeloid-specific) knockout mice and BM transplantations, we demonstrated that the systemic metabolic condition of LPI (mainly hypoargininemia and/or hyperammonemia) leads to defective erythropoiesis and altered RBCs, prompting thus increased erythrophagocytosis ultimately leading to highly iron loaded RPMs and BMMs and hyperferritinemia. Mechanistically, depletion of extracellular arginine and increased plasma amonia levels, as a result of Slc7a7 ablation in kidney and intestine, leads to impared development of RBCs that are more phagocyted by RPMs. Iron overload is a well known hepcidin driver through the bone morphogenic protein – 6 (BMP6) (Andriopoulos et al., 2009), hence, high levels of iron triggered hepcidin expression which in turn downregulated FPN1 in Slc7a7 macrophages, by this way altering iron recycling. Further analyses revealed that defective erythropoiesis was absent in the myeloid-specific knockout mouse. These findings connect two previously unrelated biological processes, namely defective RBC generation and macrophage iron accumulation, and implicate the LPI metabolic derangement as a key player in the hematologic complications of the disease.
Results
Global y+LAT1 ablation in adult mice results in a drastic reduction of bone marrow macrophages and red pulp macrophages
Y+LAT1 is highly expressed in epithelial cells and in some non-polarized cells such as macrophages (Pollard, 2009; Rotoli et al., 2020). To dissect the role of y+LAT1 in the immune and hematological complications of LPI, we created a y+LAT1 conditional allele (Slc7a7loxp/+) and generated a Slc7a7loxp/loxp/UBC-Cre+ inducible knockout model that expresses Cre in all the body cells in response to tamoxifen treatment. Twelve-week-old Slc7a7loxp/loxp/UBC-Cre+(Slc7a7-/-, after tamoxifen induction) mice and their control Slc7a7loxp/loxp/UBC-Cre- (Slc7a7+/+) littermates were first treated with tamoxifen for 7 days to induce Cre expression and were further kept on low-protein diet. As a result, y+LAT1 depletion led to significant reduction of spleen weight (Figure 1A). However, consistent with previous reports that citrulline administration ameliorates hypoargininemia and hyperammonemia as well as the vast majority of the defects caused by y+LAT1 ablation (Bodoy et al., 2019), 10 days of citrulline supplementation also improved spleen weight (Supplementary Figure 1A). Of note, Slc7a7 ablation led to lower levels of F4/80-positive cells in spleen and BM sections, together with a reduced number of F4/80hi cd11blo cells in vivo (Figure 1B-C), which were also recovered by citrulline administration (Supplementary Figure 1B-C). Besides, circulating monocytes levels were also decreased in y+LAT1-deficient mice, thereby indicating a loss of macrophage precursors (Figure 1D).
We then asked whether the decreased number of RPMs was associated with a reduced proliferation or increased apoptosis. To this end, we assessed Ki67 and caspase-3 expression in spleen sections. Indeed, Ki67 was reduced in the red pulp area of Slc7a7-/- mice, where RPMs reside (Figure 1E). Moreover, Slc7a7-/- RPMs expressed higher levels of active caspase-3 compared with those of control mice (Figure 1F). Taken together, our data suggests that y+LAT1 participates in the homeostasis of BMMs and RPMs, which might be caused by both increased apoptosis and decreased levels of proliferation.
y+LAT1 deficiency in myeloid cell line does not reproduce conditional knockout mice deficiencies
We then asked whether the lack of Slc7a7 expression in macrophages was key for the defects previously observed (Figure 1 A,C). For that purpose, we generated a myeloid cell-specific knockout mouse (Slc7a7loxp/lox LysM-Cre+; Slc7a7LysM-/-) in which Cre expression is specifically restricted to the myeloid cell lineage. Likewise, the inducible knockout mice, Slc7a7LysM, showed a reduced expression of Slc7a7 in RPMs, AMs and BMDMs (Supp Figure 2A-B). Nevertheless, contrary to the inducible model Slc7a7-/- (Bodoy et al., 2019), Slc7a7LysM mice did not present reduction of y+LAT1 expression in kidney, hypoargininemia or urea cycle dysfunction (e.g., increased orotic acid in urine), nor did they show a reduction in spleen size or body weight or a decrease in RPMs and BMMs numbers in vivo (Supp Figure 2B-G). These findings rule out the possibility that the defect comes from the lack of y+LAT1 expression in the myeloid cell line and embraces the possibility that systemic reduction of arginine along with other unbalanced amino acids and metabolites (LPI systemic metabolic condition) might underlie the observed decrease in macrophage generation and survival in total knockout mice.
Slc7a7-/- but not Slc7a7LysM red pulp macrophages, have a dysfunctional iron metabolism
Since one of the main functions of RPMs is to regulate erythrocyte degradation and iron delivery for erythropoiesis (Kohyama et al., 2009), we next focused on whether iron metabolism was compromised in both mouse models. Slc7a7-/- mice showed a dramatic iron accumulation in the BM and spleen. In contrast, abnormal iron accumulation was not visible in Slc7a7LysM animals (Figure 2A), thus, further confirming the premise that y+LAT1 deficiency in the myeloid cell line does not cause macrophage dysfunction. Citrulline treatment improved iron accumulation in the spleen and BM in Slc7a7-/- mouse (Figure 2A). In line with this finding, iron content in the liver and spleen tissue were higher in the Slc7a7-/- mice compared to its control littermates and were rescued by citrulline supplementation (Figure 2B).
We next addressed whether iron accumulation in tissues would also be reflected into higher serum ferritin levels (Cohen et al., 2010), a trait usually reported in LPI patients (Ogier de Baulny et al., 2012). Hyperferritinemia was found in the Slc7a7 knockout mice (Figure 2C), and, as expected, the impairment was reversed by citrulline administration and no differences were observed in the Slc7a7LysM mouse model (Figure 2C). To study whether hyperferritinemia was associated with increased inflammation (Kawasumi et al., 2014; Rosário et al., 2013), we examined IL6 plasma levels. Strikingly, control and Slc7a7-/- mice showed similar levels of IL6 (Supp Figure 3A), excluding thereby systemic inflammation as a plausible cause of increased ferritinemia or macrophage iron retention (Theurl et al., 2016). Given that y+LAT1 deletion resulted in a significant reduction in the number of RPMs and BMMs, we tested whether the observed iron accumulation was directly linked to macrophages. Of note, within the spleen, iron accumulation was specifically located in the resident RPMs (Figure 2D).
As a whole, these results indicate that the systemic metabolic conditions of LPI cause detrimental effects on RPM homeostasis but Slc7a7 expression in macrophages is not required for the iron accumulation in tissue.
Slc7a7-/- mouse model show reduced expression of FPN1 in macrophages
To gain insight into how iron is accumulated in macrophages, we analyzed the expression of FPN1, the only known iron exporter, and its relationship to circulating concentrations of hepcidin, the major iron-regulatory hormone that interacts directly with FPN1 triggering its degradation (Nemeth et al., 2004). Of note, hepcidin plasma levels were increased in Slc7a7-/- mice (Figure 3A) which were paralleled by increased expression levels of liver hepcidin (Hamp1) in Slc7a7-/- mice as compared to control littermates (Figure 3B). Hepcidin levels can be regulated at the transcriptional levels by several factors such as inflammation and hepatocyte iron deposits(Sebastiani et al., 2016) . In this regard, microarray data on sorted RPMs from Slc7a7+/+ and Slc7a7-/- mouse showed decreased expression of inflammatory-related pathways. Moreover, as indicated above, plasma interleukin 6 levels showed similar levels between both genotypes, thus ruling out the possibility of inflammation as a plausible cause for increased hepcidin levels (Supplementary Figure 3A-B). Nevertheless, enhanced Pearl’s Prussian blue staining revealed that Slc7a7-/- liver sections showed specific localization of iron deposits in hepatocytes (Figure 3C), suggesting thus iron accumulation as the main cause for increased hepcidin levels. The increased number of iron deposits was accompanied by a significant increase of the BMP6 (Figure 3D), a protein known to interact with hemojuvelin to further (Core et al., 2014), trigger Hamp1 transcription and expression (Andriopoulos et al., 2009; JL et al., 2006). Thus, further supporting the premise that hepatocyte liver accumulation triggers increased hepcidin plasma levels.
Flow cytometry analysis revealed a decreased number of FPN1-positive RPMs in Slc7a7-/- compared to wildtype mice (Figure 3E). Moreover, decreased FPN1 expression in RPMs was also confirmed by western blot, where Slc7a7-/- mice show reduced levels of FPN1 in total spleen (Figure 3F).
Together, our results indicate that Slc7a7 expression is necessary to maintain proper FPN1 expression and that its imbalance leads to an aberrant iron retention in resident macrophages.
Slc7a7-/- mice show exacerbated erythrophagocytosis
Our findings that Slc7a7-/- RPM were loaded with iron together with the imbalance on iron metabolism suggests that RPM function is impared in Slc7a7-/- knockout mouse. In that sense,gene expression analysis of sorted RPMs from Slc7a7+/+ and Slc7a7-/- animals revealed altered expression of key RPM-associated genes (Figure 4A) (Haldar et al., 2014; Kohyama et al., 2009) further supporting an impaired RPM functioning and disrupted iron handling. Moreover, RPMs are a highly specialized erythrophagocytic cell type in which several genes have been proposed as master regulators of iron homeostasis and RBC clearance (Kohyama et al., 2009). In this regard, these animals showed increased expression of SpiC, the master regulator of RPM differentiation, Msr1, the macrophage scavenger receptor 1, and the hemoglobin scavenger receptor CD163 in Slc7a7-/- mice, while Il1b gene expression, a proinflammatory cytokine linked to erythrophagocytosis (A-Gonzalez et al., 2017; Guo et al., 2019; Kohyama et al., 2009; Oexle et al., 2003), was significantly reduced (Figure 4B) (Moestrup and Møller, 2004). This expression pattern suggested an increased erythrophagocytosis activity in Slc7a7-/- mice. To examine this effect in further detail, we measured the ertyrhophagocytosis ratio of BMDMs from wildtype and knockout mouse. Strikingly, when Slc7a7-/- and control RBCs were co-incubated with BMDMs from Slc7a7-/- or control mice, Slc7a7-/- erythrocytes were preferentially engulfed by both macrophages (Figure 4C). Hence, ex vivo experiments confirmed a significantly higher phagocytosis rate of RBCs derived from Slc7a7-/- deficient mice compared to RBCs from control animals. In addition, RBCs from Slc7a7-/- and control animals treated with citrulline were equally engulfed by both macrophages, demonstrating again the rescuing effect of citrulline (data not shown). These results indicate that the observed iron accumulation in macrophages is due to a defect on the RBCs rather than an alteration on the RPMs functioning.
y+LAT1 depletion results in defective erythropoiesis
To understand the mechanisms underlying the defect in erythrocytes that results in increased erythrophagocytosis (Figure 4B), we performed a hematological analysis of Slc7a7-/- animals and its control littermates. Slc7a7-/- erythrocytes had a reduced mean corpuscular volume (MCV) and hemoglobin (MCH) as well as a decreased mean platelet volume (MPV), while no differences were found in Slc7a7LysM mice (Figure 5A-F). Interestingly, erythropoietic progenitors analyzed by flow cytometry showed that Slc7a7-/- mice have a severe reduction in erythrocyte precursors (Figure 5G). The dramatic decrease in erythroid precursors implies that LPI systemic metabolic condition caused by global Slc7a7 ablation compromises RBC generation. In fact, erythroid precursors of Slc7a7LysM mice were not affected (Figure 5G).
EPO is a secreted hormone responsible for stimulating RBC production and survival. Specifically, EPO has been described to stimulate RBC generation at the proerythroblast stage (Hattangadi et al., 2011). We therefore measured EPO plasma levels and found that, indeed, EPO levels were significantly decreased in Slc7a7-/- mice (Figure 5H). As a whole, these results indicate that y+LAT1 plays a specific role in erythroid development at multiple proerythroblast stages, possibly orchestrated by the metabolic defects caused by Slc7a7 ablation.
Bone marrow transplant improves iron accumulation but not the metabolic complications and erythropoietic defects
Since immature erythroid precursors were dramatically reduced in Slc7a7-/- mice (Figure 5G), RBCs had an altered MCV and MCH (Figure 5D-E), and Slc7a7LysM animals did not show any erythropoietic defects (Figure 5A-G), we speculated that the observed alterations in the Slc7a7-/- background could be due to the modified systemic microenvironment rather than a cell-autonomous defect in the BM. For this purpose, BM cells isolated from Slc7a7+/+ mice (CD45.1) were harvested and transplanted into lethally irradiated Slc7a7-/- mice (CD45.2); conversely, BM cells isolated from Slc7a7-
/- mice (CD45.2) were harvested and transplanted into lethally irradiated Slc7a7+/+ mice (CD45.1). Five weeks after transplantation, mice were placed on a tamoxifen diet for 7 days and then further fed a low-protein diet for 10 days prior to the sacrifice day (Figure 6A). In this setting, BM transplantation did not improve the body and spleen weight of the Slc7a7-/- animals (CD45.2) transplanted with Slc7a7+/+ BM (CD45.1) (Figure 6B-C). Analysis of peripheral blood revealed that arginine plasma levels were diminished and orotic acid levels in urine were high as a consequence of the urea cycle dysfunction (Figure 6D-E). Thus, as expected, BM transplant did not affect the main metabolic complications of LPI. Interestingly, Slc7a7-/- mice receiving Slc7a7+/+ BM, did not present iron accumulation in the spleen nor differences in the MCH (Figure 6F, J) and exhibited a tendency towards lower plasma ferritin levels (Figure 6G). This finding thus indicates that BM transplant can specifically rescue iron metabolism defects. Conversely, Slc7a7-/- transplanted mice exhibited a vast decrease in erythroid progenitors I-IV compartments together with reduced MCV (Figure 6H), reaffirming the profound systemic effect of y+LAT1 deficiency on RBC generation and homeostasis.
Discussion
The y+LAT1 transporter is a cationic/neutral amino acid exchanger that provides arginine for different processes in the organism. Proper arginine availability is essential for a normal urea cycle (Morris, 2002). The defiency of y+LAT1 transporter in LPI condition, causes systemic hypoargininemia, which due to the shortage of urea cycle intermediates, results in hyperammonemia (Bodoy et al., 2019). In our mouse model of LPI, as shown here, this metabolic derrangement causes erythropoiesis failure, whereas y+LAT1 expression in macrophages is not enough to cause impared development of RBCs in the Slc7a7LysM mouse model. In this regard, little is known about the role of amino acid availability in erythropoiesis. The impact of L-arginine in erythrocytes is highlighted by the fact that ablation of the arginine transporter CAT1 in mouse results in perinatal death and anemia (Perkins et al., 1997) and that L-arginine-mediated CAT1 transport participates in erythrocyte differentiation and proliferation in vitro (Shima et al., 2006). Moreover, the L-arginine catalytic enzyme endothelial nitric oxide synthase (eNOS) can be found in erythrocytes, where parasite-arginine deprivation decreases deformability of these cells as a result of reduced NO production (Cobbold et al., 2016). Here we show that upon ablation of Slc7a7 in all cells, erythrocytes present reduced mean corpuscular volume and mean corpuscular hemoglobine, which has also been also described in LPI patients (Al-Qattan et al., 2021). Since arginine is the metabolite recovered by citrulline administration, we hypothesized that hypoargininemia plays a key role in erythropoiesis and RBC size. Altered RBCs in Slc7a7-/- leads to increased erythrophagocitosis. Furthermore, bone marrow-derived macrophages from y+LAT1 knock out mice did not show an increased rate of erythrophagocytosis when exposed to control erythrocytes, further supporting that the observed abnormalities in Slc7a7-/- RPMs (i.e. increased markers of erythrophagocytosis) are primed by altered erythrocytes rather than by a cell-autonomous defect of RPMs.
RPMs prime erythrocyte degradation and iron recycling, and defects in numerous pathways can lead to iron overload in macrophages (Knutson et al., 2005). Such pathological conditions can thus lead to compromised iron metabolism and have an impact on macrophages (Ganz, 2012; Soares and Hamza, 2016). For instance, aged or damaged erythrocytes can express “eat me” signals and acutely trigger the erythrophagocytosis machinery (Luo et al., 2016; Oldenborg, 2000; Park and Kim, 2017) subsequently causing iron accumulation in RPMs (Dichtl et al., 2018) as we clearly observed in the Slc7a7-/- model. We postulate therefore that defective arginine availability is also at the basis of iron accumulation because citrulline administration recovers normal iron levels in liver, spleen and BM, and they are also not present in Slc7a7LysM tissues.
FPN1 is the only known iron exporter involved in iron efflux in macrophages. Mechanistically, FPN1 is regulated at the protein level by hepcidin, which mediates its degradation (Drakesmith et al., 2015). In this regard, the increased hepatic iron content and ferritin levels (Nemeth and Ganz, 2009) would be at the root of induced hepcidin expression as depicted by increased BMP6 liver mRNA expression. As a consequence of increased plasma hepcidin, Slc7a7-/- show decreased FPN1 expression in RPMs, which was further confirmed by both flow cytometry and western blot analysis. Presumably, in this LPI mouse model the dysfunctional RBC generation increases erythrophagocytosis by a mechanism that, at the same time, leads to overwork of the fewer RPMs and disrupted iron handling by altered FPN1 homeostasis.
Erythroblast differentiation is orchestrated mainly by EPO expression (Jelkmann, 2011; Moritz et al., 1997). Our findings that Slc7a7-/- mice show reduced EPO in circulation, highlights the importance of Slc7a7-mediated systemic arginine availability. However, while BM transplant failed to rescue the hematological alterations, it did recover macrophage iron accumulation, suggesting thus two additive mechanisms. On one hand defects in RBC maturation are caused by an unbalanced systemic metabolic environment rather than by a cell-autonomous defect. On the other hand, macrophage iron accumulation also requires Slc7a7 ablation in macrophages. Future work is needed to elucidate the link between Slc7a7-mediated systemic arginine availability and EPO production, and between macrophage y+LAT1 and FPN1 expression.
Substantial advances in the research field of LPI have been scarce during the last decade, which can be probably explained by the huge phenotypic variability found among LPI patients (Al-Qattan et al., 2021; Ogier de Baulny et al., 2012; Posey et al., 2014). In a previous work we showed that the inducible total Slc7a7-/- mouse model recapitulates the main hallmarks of the human LPI complications, such as hypoargininemia, hyperammonemia and PAP (Bodoy et al., 2019; Ogier de Baulny et al., 2012; Parto et al., 1994). In the current work we show that Slc7a7-/- mouse model has increased erytrophagocitosis, elevated serum ferritin, altered hemogram and abnormal iron retention in macrophages (Ogier de Baulny et al., 2012). Moreover, the systemic metabolic condition of LPI is an essential driver for the hematologic complications. In addition, Slc7a7-/- mice showed deficient erythropoyesis, a trait that paralels the reduced number of reticulocytes in some LPI patients, an alteration that has not been fully explored (Al-Qattan et al., 2021). Our results shifted the spotlight of increased erythrophagocytosis by abnormally functioning macrophages to altered erythrocytes.
Historically, systemic metabolic condition and immune-hematologic complications of LPI have been considered as independent entities, such is the case that some authors concerned about a potential detrimental effect of citrulline treatment in the development of immune complications (Ogier de Baulny et al., 2012). Nevertheless, more recently low argininemia was revealed as a poor prognosis factor in LPI (Mauhin et al., 2017). Now, our work upholds the premise that the handling of the metabolic derangements could prove beneficial not only for the metabolic hallmarks of the diseases but also for the reported hematologic complications of the patients.
Author Contributions
F.S., J.G., and S.B. designed and performed experiments, interpreted and analyzed data. J.G., S.B., and M.P. designed research and wrote the manuscript with input from all of the authors. J.C. performed research. G.W and M.S. provided reagents and provided intellectual input. A.O., R.A. and A.Z. provided reagents.
The authors declare no competing financial interests.
MATERIALS AND METHODS
Data and code availability
Microarray data has been deposited in a public repository and the accession numbers is GSE164827.
Animals
All animal work was approved and conducted according to guidelines established. This project (DARP n°9177) has been assessed favourably by the Institutional Animal Care and Use Committee from Parc Científic de Barcelona (IACUC-PCB) and the IACUC considers that the above-mentioned project complies with standard ethical regulations and meets the requirements of current applicable legislation (RD 53/2013 Council Directive; 2010/63/UE; Order 214/1997/GC). C57BL/6 mice were purchased from Harlan Europe. Slc7a7loxp/loxp mice were generated by Eurogentec. To generate Slc7a7-/- and Slc7a7LysM mice, Slc7a7loxp/loxp were crossed with UBC-Cre-ERT2 mice from The Jackson Laboratory and LysM-Cre provided by Dr. Ángel R. Nebreda, respectively. Male or female mice of 12 weeks old were used. Mice were housed in groups of 2-5 animals per cage and were kept under a 12 h dark-light period. Food and water were supplied ad libitum. Animals were fed a standard diet (Teklad global 14% protein rodent maintenance diet) until tamoxifen induction, which consisted of a tamoxifen diet for one week. After the induction period, animals were kept on a low-protein diet for 7-10 days, supplemented or not with 1g/l L-citrulline in drinking water. Control and Slc7a7-/- littermates on a C57Bl6/J genetic background were sacrificed at 10-12 weeks of age by cervical dislocation. Tissues were dissected and flash-frozen in liquid nitrogen for RNA, protein, and iron quantification studies. For histological analysis, mice were anesthetized with ketamine and xylazin (respectively 1 mg and 0.1 mg per 10 g of body weight, i.p., respectively) and subjected to transcranial perfusion. For hematological and biochemical studies, EDTA or heparin blood was collected from cardiac puncture. Bone marrow was flushed out from femur and tibia bones.
Mice carrying the myeloid-specific knockout of the Slc7a7 gene (LysMCre/+ Slc7a7flox/flox) were sacrificed at 12 weeks of age, and only those with more than 80% deletion of endogenous protein were used for the experiments.
Bone marrow transplantation (BMT)
Recipient mice were lethally irradiated (9.5Gy) and transplanted with 2x106 bone marrow (BM) cells by retro-orbital injection (Bennett et al., 2018). For the re-population experiments, total BM cells from either Slc7a7-/- (CD45.2) or Slc7a7+/+ (CD45.1) mice were transplanted into lethally irradiated B6 recipient mice (CD45.1 or CD45.2). As a follow-up step, five weeks after transplantation to allow whole body hematopoiesis regeneration, mice were subjected to tamoxifen diet for 7 days, and then treated with a low-protein diet for 10 days prior to the sacrifice. BM reconstitution was monitored by flow cytometry.
γ-Irradiation of mice was performed in a 137Cs-γ IBL 437C H irradiator (Shering CIS bio international) at 2.56Gy/min rate for the indicated dosage. The irradiated mice were inspected daily. Mice were given Baytril water containing antibiotics (Bayer, Shawnee Mission, JS) for at least 30 days to reduce the probability of infection from opportunistic pathogens.
Flow cytometry and cell sorting
For the analysis of splenocytes and bone marrow cells, crushed spleens and flushed BM were isolated and incubated with Fc block (anti-mouse CD16/32; Thermofisher) for 30 min on ice. Cell suspensions were stained for the expression of CD71; CD11b; CD45.1; CD45.2 (BD Biosciences); CD34 (eBiosciences); F4/80; TER119; CD106 (BioLegend) for 30 min on ice. Flow cytometry analysis was performed on Gallios (BD Boisciences). For spleen staining, crushed tissues were filtered through a 40 μM cell strainer and erythroid cells were removed by incubation with ammonium-chloride-potassium lysis buffer prior to Fc blocking. Cell sorting (purity > 90%) was carried out using a FACS Aria II (BD Biosciences). For microarray analysis, spleens were prepared as described above and stained with anti-CD106, anti-CD11b and anti-F4/80 (ThermoFisher) for purified RPMs.
Cell doublets were excluded from all analyses and, when possible, dead cells were excluded by the use of DAPI. Data analysis was carried out using FlowJoTM Software.
In vitro erythrophagocytosis assay
To prepare primary BMDMs, cells obtained from mouse femurs and tibia were cultured for 7 days in the presence of L-Cell (L929 SN) in DMEM supplemented with 10% FBS, penicillin (50 U/mL) and streptomycin (50 μg/mL). BMDMs were plated 24 hours prior to the day of the experiment. On the day of the experiment, previously seeded BMDMs were activated with lipopolysaccharide (100 ng/mL) for 2h and fresh RBCs were extracted, washed and labelled with CellVueⓇ or PKH26 following the manufacturer’s instructions. RBCs were then incubated with previously activated BMDMs for 2 minutes (10·106 RBC/1·106 BMDM) at 37°C in a 5% CO2 incubator. Macrophages were washed twice with PBS and finally incubated with an erythrolysis buffer (R&D Systems) to lyse non-ingested RBCs. Cells were then collected and analyzed by flow cytometry.
Histological sample preparation and analysis
Samples were fixed overnight at 4°C with neutral buffered formalin. After fixation, bone tissue (femur) was washed with PBS 1x and decalcified with Osteosoft® reagent for a minimum 15 days at RT. All samples were embedded in paraffin. Paraffin-embedded tissue sections (2-3 μm in thickness) were air-dried and further dried at 60 °C overnight. Bone sections were maintained at 60°C for 48 h.
For special staining, paraffin-embedded tissue sections were dewaxed and stained with Iron Stain Kit to identify iron pigment using the Dako Autostainer Plus and following the manufacturer instructions. When combining Iron staining with F4/80 IHC, iron staining was done before following the described protocols.
Prior to immunohistochemistry, sections were dewaxed and therefore epitope retrieval was performed using citrate buffer pH6 for 20 min at 121°C with an autoclave or proteinase K for 5 min at RT for anti-caspase 3 (Cell Signalling) and rat monoclonal Anti-F4/80 (eBioscience), respectively. For rabbit polyclonal anti-Ki67 (Abcam) sections were dewaxed as part of the antigen retrieval process using the low pH EnVision™ FLEX Target Retrieval Solutions (Dako, Burlington) for 20min at 97°C using a PT Link (Dako, Agilent). Quenching of endogenous peroxidase was performed by 10 min of incubation with Peroxidase-Blocking Solution at RT. Non-specific bindings were blocked using 5 % of goat normal serum or normal donkey serum mixed with 2.5 % BSA diluted in the wash buffer for 60 min at RT. The primary antibody dilutions used were 1:300, 1:100 and 1:2000, for 120 min, overnight or 60 min, respectively. The secondary antibody used was a BrightVision Poly-HRP-Anti Rabbit IgG Biotin-free, ready to use or the secondary antibody used was a Biotin-SP (long spacer) AffiniPure Donkey Anti-Rat IgG (H+L) at 1:500 (in wash buffer) for 60 min followed by amplification with Streptavidin-Peroxidase polymer at 1:1000. Antigen– antibody complexes were revealed with 3-3′-diaminobenzidine, with the same time exposure (1 min). Sections were counterstained with hematoxylin and mounted with Mounting Medium, Toluene-Free using a Dako CoverStainer. Specificity of staining was confirmed with rabbit IgG, polyclonal - Isotype control or Normal Rat IgG Control. Image acquisition. Brightfield images were acquired with a NanoZoomer-2.0 HT C9600 digital scanner (Hamamatsu) equipped with a 20X objective. All images were visualized with the NDP.view 2 U123888-01 software. All images were visualized with a gamma correction set at 1.8 in the image control panel of the NDP.view 2 U12388-01 software.
Prior to immunohistochemistry, for Ki67 sections were dewaxed as part of the antigen retrieval process using the low pH EnVision™ FLEX Target Retrieval Solutions (Dako, Burlington) for 20 min at 97°C using a PT Link (Dako – Agilent). For caspase 3 samples were dewaxed and antigen retrieval treatment was performed with citrate buffer pH6 for 20 min at 121°C with an autoclave. Quenching of endogenous peroxidase was performed by 10 min of incubation with Peroxidase-Blocking Solution (Dako REAL S2023). Rabbit polyclonal primary anti-Ki67 antibody (A. Menarini diagnostics – NCL-ki67p) was diluted 1:1000 with EnVision FLEX Antibody Diluent (K800621, Dako, Agilent) and incubated for 60 min at RT. The secondary antibody used was a BrightVision Poly-HRP-Anti Rabbit IgG Biotin-free, ready to use (Immunologic, DPVR-110HRP). Antigen–antibody complexes were revealed with 3-3′-diaminobenzidine, with the same time exposure per antibody (3 and 5 min respectively). Sections were counterstained with hematoxylin and mounted with Mounting Medium, Toluene-Free using a Dako CoverStainer.
Amino acid content
Briefly, amino acids were determined by ion exchange chromatography with ninhydrin derivatization and spectrometric detection (Biochrom 30, Chromsystems, Cambridge, UK). Plasma (300 µL) were deproteinized with sulphosalicylic acid containing L-norleucine as internal standard (final concentration 100 µmol/L). After centrifugation, 200 µL of supernatant were adjusted to pH = 2.1 with lithium hydroxide, and then, injected onto the liquid chromatograph. Urinary orotic acid was analyzed following a spectrometric procedure (458 nm), by reacting with para-dimethylaminobenzaldehyde.
Tissue iron content
Liver and spleen non-heme iron content was measured using the bathophenanthroline colorimetric method. Mouse tissues were dried at 45°C for 3 days, weighted, and digested for 48 h at 65°C in 10% TCA/10% HCl to allow deproteinization of non-heme iron. Diluted extracts were added to a 0.01% bathophenanthroline disulfonic acid, 0.1% thioglycolic, 7M sodium acetate solution and the absorbance at 535 nm was measured using a spectrophotometer Ultrospec 3100pro (Amersham Biosciences). The iron content of samples was obtained by interpolation from a standard curve and calibrated to the weight of dried material (Jd and Th, 1968; Patel et al., 2002).
Plasma measurements
ELISA kit was used to determine the IL6 (Abnova), hepcidin (Intrinsic Life Science), ferritin (Abcam) and erythropoietin (R&D Systems) proteins in fresh plasma. The procedures were done following the manufacturer’s instructions.
Microarray analysis
For gene expression analysis of RPMs, total RNA was isolated from previously purified cells using magnetic beads and the Agencourt RNA Clean XP kit (Beckman Coulter). Quality and quantity were assessed using a Bioanalyzer 2100 (Agilent Technologies, Palo Alto, CA). Library preparation and amplification were performed as described previously by (Gonzalez-Roca et al., 2010). RNA was amplified for 22 cycles and purified using PureLink Quick PCR Purification kit (Invitrogen) in the Genomic Facility of IRB Barcelona.
Pre-processing of microarray data
Microarray datasets were processed separately using R (R Core, 2019) packages affy (Gautier et al., 2004) and affyPLM (Bolstad et al., 2005) from Bioconductor (Gentleman et al., 2004). Raw cell files data were processed using RMA (Irizarry, 2003) and annotated using the information available on the Affymetrix – Thermofisher web page. Standard quality controls were performed in order to identify abnormal samples regarding: a) spatial artefacts in the hybridization process (scan images and pseudo-images from probe level models); b) intensity dependences of differences between chips (MvA plots); c) RNA quality (RNA digest plot); d) global intensity levels (boxplot of perfect match log-intensity distributions before and after normalization and RLE plots); and e) anomalous intensity profile compared to the rest of the samples (NUSE plots, Principal Component Analysis).
Differential expression
A differential expression analysis was performed for Slc7a7+/+ and Slc7a7-/- comparisons using a linear model with empirical shrinkage (Smyth, 2004) as implemented in Limma R package (Ritchie et al., 2015). This model included the batch of scanning for statistical control. Adjustment by multiple comparisons was performed using the Benjamini-Hochberg method (Benjamini and Hochberg, 1995).
Biological enrichment analysis
Genes quantified in the microarray experiment were annotated according to the Broad Hallmark (Liberzon et al., 2015) gene sets collection. Broad Hallmark sets were translated to mouse homologous genes using the R package biomaRt (Durinck et al., 2009).
Functional enrichment analyses were performed using a modification of ROAST (Wu et al., 2010), a rotation-based approach implemented in the R package limma (Ritchie et al., 2015) that is especially suitable for small size experiments. Such modifications were implemented to accommodate in the ROAST algorithm the statistical re-standardization proposed in (Efron and Tibshirani, 2007), which enables its use for competitive testing (Goeman and Bühlmann, 2007). The MaxMean (Efron and Tibshirani, 2007) statistic was used for testing geneset enrichment of Broad Hallmark. For each gene, the most variable probeset within each gene was used in these analyses (median absolute deviation).
The results of these analyses were adjusted by multiple comparisons using the Benjamini-Hochberg False Discovery Rate method (Benjamini and Hochberg, 1995).
Clustering and visualization
Gene expression of selected genes was graphically represented in a heatmap with the heatmap R package, using a blue to red gradation, where red indicated the highest expression and blue corresponded to the lowest expression values. Previously, the expression data were summarized to the gene level using the most variable probeset mapping to the same gene (median absolute deviation), and expression values were centered and scaled gene-wise. Genes and samples were clustered using the Ward agglomeration method and the correlation and Euclidean distances, respectively. To gain clarity in the graphic, the most extreme values were truncated to -1.5 and 1.5. All analyses were carried out using R and Bioconductor.
RNA extraction and quantitative real-time PCR
Mice were killed by cervical dislocation, and tissues were immediately frozen for RNA isolation. Total mRNA was extracted from BMDMs or AMs using the Rneasy Total RNA Isolation kit (Qiagen, Alameda, CA, USA), following the manufacturer’s instructions. RNA concentrations were measured with Nanodrop ND-1000 (ThermoFisher Scientific). Reverse transcription was performed with total RNA (2 ng) using the qScript cDNA SuperMix (Quantabio) following the manufacturer’s instructions. PCRs were performed using the ABI Prism 7900 HT real-time PCR machine (Applied Biosystems, USA) and the SYBR® Green PCR Master Mix. Gene expression levels were normalized with β-actin as housekeeping genes. Primers used are listed in Supplementary table 1.
Protein isolation and western blot
Membrane proteins from cell cultures or tissues were extracted with Lysis buffer (25 mM Hepes, 4 mM EDTA, 250 mM Sucrose) containing protease inhibitor (1:000; Protease Inhibitor Cocktail Set III, EDTA-Free, Calbiochem). Briefly, tissues were lysed using the Tissue Lyser (Mini-beadbeater-16, Biospecproducts) and further centrifuged at 10000 g for 10 minutes at 4°C. After centrifuging, the supernatant was centrifuged again on an ultracentrifuge at 55000 rpm for 1 hour at 4°C. Finally, protein concentration was determined using Pierce BCA Protein Assay Kit (ThermoFisher Scientific). Membrane proteins were resolved in 10% acrylamide gels for SDS-PAGE and transferred to Immobilon membranes (Millipore). The following antibodies were used: polyclonal rabbit anti-y+LAT1 was used at 1:750 dilution with 5% non-fat dried milk in PBS Tween-20 (0.1%) (Bodoy et al., 2019); rabbit anti-FPN1 was used at 1:250 dilution with 5% non-fat dried milk in TBS Tween-20 (0.1%) (Nairz et al., 2013). Antibody binding was then detected using appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies (1:1000 dilution). Proteins were detected by the enhanced chemiluminescence method (GE Healthcare Life Sciences) and quantified by scanning densitometry.
Primary bone marrow macrophages (BMDMs) cell culture
BM cells from 12-week-old mice (either female or male) were flushed from mice femurs and tibias. The cell suspension was lysed for 5 min in ACK lysis buffer at RT and then washed, resuspended, and cultured for 7 days in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 50 U/mL penicillin, 50 μg/mL streptomycin and 50 ng/mL of recombinant M-CSF (Peprotech) or 30% of L-Cell (L929 supernatant (SN)) media. Six days after the seeding, cells were harvested and re-seeded with the specific conditioned media for 24 hours. To deplete arginine, arginine-free media was used (DMEM for SILAC, ThermoFisher).
QUANTIFICATION AND STATISTICAL ANALYSIS
Data were analyzed using GraphPad Prism Version 8 software. Statistical analysis was performed using the Student’s t test and one-and two-way ANOVA as specified in each figure legend.
Acknowledgements
This work was supported by grants from the Spanish Ministry of Science and Innovation (grant SAF2015-64869-R-FEDER and RTI2018-094211-B-100), Ramon Areces Fundation (I.O.F.R.ARECES) and the Generalitat de Catalunya (grant 2017 SGR 961). Grant RTI2018-101735-B-I00 from the Spanish Ministry of Science and Innovation to MS. The EMBO Short Term Fellowship Program facilitated the collaboration between international groups. We are also grateful to Dr. Angel Nebreda at the IRB Barcelona, Spain, for LysM-Cre transgenic mice; Dra Anna Bigas from IMIM, Spain, for CD45.1 mice. We thank Jorge Seco and Vanessa Hernández for technical assistance and help with the experimental animals, and the Daniel Bravo Foundation for support with amino acid determination.