P1PK blood group antigens in birds: implications for veterinary blood transfusion, human and pet health Authors: Anna Bereznicka, Maria Duk, Katarzyna Kapczynska, Anna Modlinska, Tomasz Piasecki, Paweł Link-Lenczowski, Małgorzata Heidorn-Czarna, Pawel Pasikowski, Marcin Czerwinski, Radoslaw Kaczmarek Published: 2025 ABSTRACT BackgroundThe P1PK blood group system in humans, composed of three glycosphingolipid antigens, plays a significant role in transfusion medicine. These antigens are synthesized by the enzyme α1,4-galactosyltransferase, encoded by the A4GALT gene, and also act as receptors for Shiga toxins, which are key virulence factors produced by pathogenic strains of Escherichia coli. Birds are known to be resistant to the effects of Shiga toxins, yet the molecular basis of this resistance has remained unclear. This study aimed to investigate the presence, expression, and functional role of P1PK antigens in birds, particularly in relation to their exposure to human-associated bacteria.ResultsBirds residing in urban environments, such as pigeons (Columba livia) and crows (Corvus cornix), express higher levels of Shiga toxin receptors on their red blood cells. These levels are higher compared to species living in more isolated areas, such as woodcocks (Scolopax rusticola). This pattern may suggest an evolutionary adaptation to human environments. We analyzed RBC glycan expression in 52 birds representing 10 species using flow cytometry, mass spectrometry, and HPTLC to assess interspecies variation in P1PK-related antigens. We propose that the increased expression of these antigens in urban birds functions as a decoy mechanism, reducing the toxicity of Shiga toxins by sequestering them before they can cause cellular damage. The glycomic analysis identified novel glycosphingolipids and glycoproteins with terminal galactose residues that are specifically recognized by the B subunit of Shiga toxin. In addition, we found that a cellular protein, RAB11FIP5, may influence the intracellular trafficking of Shiga toxins, adding another layer of protection in birds.ConclusionsOur data suggests that certain birds, particularly those residing in human-influenced environments, possess mechanisms that enable them to partially tolerate or neutralize Shiga toxins. These findings provide a basis for comparative studies of avian red blood cell glycan biology and offer valuable context for future research in veterinary transfusion science and immunology.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12917-025-05209-6. Background Since the discovery of the ABO blood group system at the beginning of the XX century by Karl Landsteiner [1], multiple studies have been conducted about the role of blood groups in health and disease [2]. According to the International Society of Blood Transfusion (ISBT), there are 49 human blood group systems, the most well-known of which are ABO and Rh (ISBT, https://www.isbtweb.org/isbt-working-parties/rcibgt/blood-group-allele-tables.html). Blood group antigens are genetically determined polymorphic traits that can be recognized by antibodies. The epitopes comprise protein or glycan moieties of glycoproteins or glycolipids. In transfusion medicine, blood group incompatibility between individuals may lead to life-threatening severe post-transfusion reactions. Blood groups are also crucial during pregnancy and childbirth. For example, the Rh blood group incompatibility between mother and newborn may lead to hemolytic disease of the newborn (HDN, Rh disease) or miscarriage [3]. Blood group antigens play many essential roles beyond transfusion medicine. They serve as receptors for pathogens, facilitate intracellular uptake and adhesion, and participate in the binding of the complement components or signal transduction. These processes may ultimately trigger host immune responses [4]. They can also act as intracellular cargo transporters. One example of the importance of blood group in host–pathogen interaction is the predominance of the O blood group in malaria-endemic areas, which results from evolutionary pressure to develop resistance against Plasmodium sp. [5]. The P1PK blood group system (ISBT no. 3) consists of three glycosphingolipid antigens: Pk and P1 with terminal sugar structures Galα1 → 4Galβ1, and NOR with terminal Galα1 → 4GalNAcβ1. The P1 antigen is synthesized in the neolacto series pathway, while Pk and NOR belong to the globo series [6]. The Pk antigen is present in red blood cells, lymphocytes, smooth muscle cells, and endothelium of most individuals except those with the p (null) phenotype, who produce none of the P1PK antigens. The lack of P1 underlies the P2 phenotype [6]. Approximately 75% of Caucasians, 20% of Asians, and 94% of Africans possess the P1 antigen in their red blood cells, resulting in the P1 phenotype. The biosynthesis of P1PK blood group antigens in humans is catalyzed by α1,4-galactosyltransferase (Gb3/CD77 synthase), encoded by A4GALT. This enzyme transfers galactose from UDP-Gal to acceptor substrates bearing terminal galactose. A previously described mutation (p.Q211E) alters its substrate specificity, allowing the formation of NOR antigen structures [7]. The substrate specificity of Gb3/CD77 synthase appears broader than initially assumed, as it can catalyze galactosylation of both glycolipids and glycoproteins, contributing to the formation of N-glycans bearing the P1 epitope [8]. P1PK blood group antigens are upregulated in ovarian cancer [9, 10], colon cancer [11], and gastric adenocarcinoma [12]. Little is known about blood group systems in birds, particularly those found in the wild. Most research on avian blood typing has focused on poultry and was conducted mainly by Briles and co-workers in the 1940s-1960s [13, 14, 14–17]. To date, 48 blood group systems have been identified in the chicken (Gallus gallus). Highly polymorphic alloantigen B was assigned as the chicken major histocompatibility complex (MHC) [18]. Among wild birds, a few chicken-like alloantigens were identified. Most published results used only one of the two typing methods, including avian interspecies agglutination (most commonly involving turkey, quail, duck, and chicken) and Ouchterlony tests (also with human or sheep antibodies or sera) [19, 20]. Studies on blood groups of wild or domestic avian species (excluding poultry) were conducted only on species belonging to Passerinformes (94 species of passerines, 18 specimens of song sparrows, and 24 species of fringillids) [21] and song sparrows from the San Francisco Bay (over 200 specimens) [22]. The authors evaluated the presence of RBC antigens belonging to the human blood groups ABO, Rh, Kell, Duffy, and Kidd. Notably, all the avian blood samples studied exhibited similar patterns. A comparison of results revealed that the frequency of blood types in birds is connected to those in humans, with the pattern 0 > AB > A > B. However, blood group 'B' had the lowest percentage or was not identified. Previous studies on avian blood transfusion, although limited, have demonstrated both the feasibility and challenges of allogenic RBC administration in birds. Early work by Degernes and co-workers [23] and later reports by Cital and co-workers [24] described transfusion outcomes in pigeons, psittacines, and raptors, highlighting issues such as short RBC survival, alloantibody formation, and species-specific immunohematological differences. These foundational observations support the need for better antigenic characterization, including P1PK-related glycans, across diverse bird species. The novel sequencing methods and bioinformatic tools showed that birds do not have any homologs of the human ABO blood group system [25]. Considering the subsequent discoveries, it became clear that blood typing in modern birds (Neoaves), proposed by Norris and other scientists in the last forty years of the twentieth century, was insufficiently precise in light of the ongoing state. On the other hand, these preliminary studies lay the groundwork for further research on interspecific interactions between humans and birds. The P1PK blood group system poses distinct challenges and offers new possibilities in veterinary practice, especially when it comes to bird health and their interactions with humans. The first studies suggesting the presence of similar human P1/P1-like antigens in wild birds were published in the 1970s. It was found that P2 individuals breeding pigeons may have high levels of anti-P1 antibodies, possibly induced by inhalation of pigeon-derived particles containing the P1 antigen [26], thus implicating immunization against P1 in the pathogenesis of pulmonary and cardiac dysfunctions that are common in pigeon breeders, such as chronic bronchitis, chronic obstructive pulmonary disease, congenital heart disease, asthma and allergies [26–31]. In addition, research focused on identifying P1 glycotope-bearing glycoproteins in other pigeon species [32–34]. Thus, wild birds produced structures similar to human P1 antigens, which evoked a humoral response by the human immune system. Pet bird owners keep multiple species of parrots, canaries, finches, fancy, and racing pigeons, leading to reciprocal immunization and pathogen transmission. Human blood group antigens may induce alloantibodies in birds, and conversely, avian antigens (e.g., pigeon P1-bearing glycoproteins) may immunize humans, leading to health sequelae. Birds can also serve as reservoirs of pathogens. Many pathogens that spread through birds can be transmitted to farm animals, pets, or humans. Understanding the complexity of avian blood group systems may help us better comprehend pathogen transmission and identify new preventive measures. In this article, we describe the presence of human P1PK blood group system counterparts in several bird species. We identify intraspecies differences and show that pigeon P1-bearing glycoproteins are not responsible for the agglutination of human RBCs. We also describe the N- and O-glycome of avian RBC membranes and present data on the novel glycosphingolipid and glycoprotein with a terminal Galα1 → 4Gal structure recognized by subunit 1B of Shiga toxin, shedding new light on why birds are resistant to Shiga toxins. Materials and methods Flow cytofluorometry All procedures complied with the Animal Welfare Act of the Republic of Poland, dated December 15, 2015. No protected or private land was used in the study. The species evaluated in the study are not protected under Polish law. All birds were privately owned, and their owners were informed of the study's aims and provided formal informed consent. Blood samples were collected during standard clinical procedures following good veterinary practice. The blood samples were primarily used for diagnostic purposes, with the leftovers reserved for experiments. Thus, per the ruling of the Local Ethics Committee for Animal Experimentation in Wrocław, operating under the auspices of the Hirszfeld Institute of Immunology and Experimental Therapy from September. As of April 4, 2019, and endorsed by the second one, operating under the auspices of Wrocław University of Environmental and Life Sciences (as of September 2025), such studies do not require ethics committee approval. Whole blood (1 mL) was collected from the right jugular vein or the cubital vein of patients at the Department of Epizootiology and Clinic of Birds and Exotic Animals, Wroclaw University of Environmental and Life Sciences, by veterinary doctors. Heparin-coated vials (Baxter Healthcare Corporation, USA) were used to separate red blood cells. Alternatively, whole blood was collected without anticoagulant and centrifuged to obtain sera. All samples were stored at 4 °C. In total, 52 birds of 10 species were evaluated (listed in Suppl. Table 1). Given the practical and ethical constraints associated with privately owned and free-living birds, the sample size (n = 52 across 10 species) reflects the maximum number of individuals available during the study period without requiring additional handling or invasive procedures. Comparable sample sizes are frequently reported in avian immunohematology research, where opportunistic sampling is standard. The dataset was therefore appropriate for the primary objective to assess interspecies differences in P1PK-related antigen expression rather than for powered statistical comparisons within individual species. Red blood cells (RBCs) from healthy birds were evaluated by flow cytofluorometry using FACSCalibur (BD Biosciences), mouse monoclonal anti-P1 antibody (dilution 1:2,000), and Shiga toxin subunit 1B (Stx1B, final concentration 1 µg/ml) as described previously [35]. The number of events analyzed was 10,000/gated cell population, and data analyses were carried out using FlowJo™ Software 10.7.1_CL [36]. Agglutination tests RBCs were washed three times with cold PBS and centrifuged for 3 min. 400 × g (Eppendorf Centrifuge 5415 R, Eppendorf) at 40C. Digestion with papain was performed as follows: 20 µL of packed RBCs were mixed 1:1 (v/v) with papain solution and incubated for 40 min. at 370C. Then, RBCs were washed 5 times with cold PBS, pH 7.4. The agglutination reactions were carried out in 96-well U-shaped polystyrene plates (NUNC, Thermo Fisher Scientific, USA). 2% RBC suspensions were mixed 1:1 (v/v) with an antibody solution or appropriately diluted sera. Agglutination was evaluated after one hour or overnight incubation at 40C. For interspecies screening, RBCS were not treated with papain. For the agglutination inhibition reaction, serial dilutions of pigeon egg ovomucoid solution containing the P1 antigen, glycophorin A (isolated from A or O blood group human RBCs), ovarian cyst (from B blood group female individual) and BSA (negative control) with the initial concentration of 200 µg/ml or 50 µg/ml (for glycophorin A of blood group O) were used. These solutions were incubated for 1 h at RT with shaking with 100X diluted pigeon serum, and then. a 2% solution of human RBCs of various phenotypes was added to the inhibitory factor-serum solution in a 1:1 (v/v) ratio. Agglutination or inhibitory effect was determined after one hour of incubation at 40C. Before every agglutination inhibition reaction, the optimum dilution of avian serum (the last with noticeable agglutination) was determined. The initial pigeon egg ovomucoid solution contained a P1 antigen concentration of 500 µg/ml for interspecies interaction screening. Detection of anti-GPA and anti-P1 antibodies in avian serum The detection of antibodies recognizing glycophorin A and P1 in the serum of wild birds (Supp. Table 2) was performed using ELISA. 96-well plates coated with glycophorin A or P1-PAA conjugate at a final concentration of 2 µg/100 µl were incubated sequentially with blocking solution (5% BSA in PBS, pH 7.4), avian serum (100-fold diluted in PBS, pH 7.4) and HRP-conjugated anti-bird IgG (1000-fold diluted in 1% BSA in PBS). Each incubation was performed for 1 h at RT. Dissolved OPD (o-phenyldiamine) reagent in a mixture of methanol and 0.1 M citrate buffer, pH 4.5, with the addition of 10 µl H2O2, was used to develop the reaction. Readings were taken at λ = 490 nm using an EnSpire 2300 Multilabel Reader (PerkinElmer, Waltham, MA, USA). However, we observed a relatively high nonspecific binding of anti-IgG antibodies to glycophorin A (Kneg_p, where pigeon serum was not used). Protein lysates, SDS-PAGE and Western blotting Protein lysates from 100 μl washed with PBS RBCs were prepared using CelLytic M® reagent (Sigma, USA) according to the manufacturer's protocol as described in Bereznicka et al. [35]. The Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, USA) was used to measure the total protein concentration in the samples. The SDS-PAGE electrophoresis was prepared according to the Laemmli method [37], and the gels were visualized with Coomassie Brilliant Blue R-250 (Roth, Germany), or transferred to the nitrocellulose membrane (Roth, Germany) as previously published [35]. 2-D electrophoresis Lyophilized pigeons RBC membranes (60 mg) were suspended in 500 μl lysis buffer (250 mM Tris–HCl pH 8.0, 50 mM NaCl, and 1X proteinases inhibitor cocktail) (Thermo Fisher Scientific, USA), passed through a 22G × 1 ½ needle 20 times, and centrifuged for 15 min with a maximum speed (15,000 × g) of 40C. The supernatant (protein lysate) was collected for additional purification before isoelectric focusing. (IEF). The protein lysate (100 μl) was purified before the first dimension run of 2D electrophoresis using ReadyPrep™ 2-D cleanup kit (BioRad, Germany), according to the protocol provided by the manufacturer. The dried precipitate was dissolved in 125 μl of 2-D sample solution ((7 M urea, 2 M thiourea, 4%CHAPS, 40 mM DTT, 0.2%(w/v) ampholytes (pH 3–10)) and applied onto IPG strips (7 cm, pH 3–10 NL R, BioRad, Germany) for rehydration (12 h in RT). The first-dimension IEF was carried out with the IEF100 system (Hoefer Inc., Holliston, USA) according to the manufacturer's protocol (Table 1). Before the second-dimension, the IPG strips were washed in the equilibration buffer ((equilibrated at RT for 15 min in 6 M urea, 30% (w/v) glycerol, 2% (w/v) SDS in 0.05 M Tris–HCl pH 8.8)) and incubated for 10 min in RT with shaking in the equilibration buffer with DTT (100 mg/10 ml, buffer 1) and iodoacetamide (400 mg/10 ml, buffer 2), respectively. After the last incubation, IPG strips were washed with 1X SDS PAGE running buffer (BioRad, Germany). Second-dimension was obtained by SDS-PAGE using 10% polyacrylamide resolving gels. The PageRuler Prestained Protein Ladder (Thermo Fisher Scientific, USA) was used as protein mass standards. ROTI ® CBB colloidal solution was used for the visualization of protein spots.Table 1First-dimension IEF programmed protocolStepVoltage [V]Time [hh: mm: ss]Gradient5000:30:00Gradient10000:30:00Gradient60000:30:00Constant60008:00:00Constant10001:1000:00 Table 1First-dimension IEF programmed protocolStepVoltage [V]Time [hh: mm: ss]Gradient5000:30:00Gradient10000:30:00Gradient60000:30:00Constant60008:00:00Constant10001:1000:00 First-dimension IEF programmed protocol LC-MS-based identification of proteins recognized by Stx1B In-gel digestion In-gel digestion of protein spots obtained in Coomassie-stained two-dimensional gel electrophoresis was performed according to the procedure of Shevchenko et al. [38]. The spots were excised from the polyacrylamide gel, cut into fragments of approximately. 1 × 1 mm, and destained to remove excess Coomassie dye using a mixture of 100 mM NH4HCO3:acetonitrile (1:1, v/v, SIGMA: Merck). The gel spots were then dehydrated by incubating twice in acetonitrile. Reduction of disulfide bridges was performed by adding 10 mM dithiothreitol (SIGMA) in 100 mM NH4HCO3 and incubating for 30 min at 56 °C. The gel fragments were then dehydrated again. Alkylation of the cysteine sulfhydryl groups was carried out by adding a 55 mM solution of iodoacetamide (SIGMA) in 100 mM NH4HCO3 to the gel fragments and incubating for 20 min in the dark at room temperature. Gel digestion of proteins was performed using 13 ng/µl trypsin solution (Promega) in 10 mM NH4HCO3 with 10% acetonitrile at 37 °C overnight. After overnight incubation, the solutions underwent additional extraction using 70% acetonitrile in water. Both solutions were then combined. LC–MS grade reagents and MiliQ water were used in each experiment step. LC-MS analysis Peptide samples from in-gel digestion were separated by liquid chromatography using a Dionex Ultimate 3000 RSLCnano (Thermo Scientific) and analyzed using MS/MS on an Orbitrap Exploris™ 480 mass spectrometer (Thermo Scientific). The following reagents and parameters were used: PepMap C18 trap cartridge (Thermo Scientific); Acclaim PepMap column, 75 µm × 25 cm (Thermo Scientific), 5 µm resin; eluent A: water, 0.1% formic acid (SIGMA); eluent B: acetonitrile: water (80:10), 0.1% formic acid (SIGMA); gradient: 5–45% eluent B, time: 40 min; eluent flow rate: 300 nl/min, injection 5 µl. LC–MS grade solvents and MiliQ water were used for eluent preparation. Mass spectra were collected in the positive ion mode (380–1500 m/z). Data analysis and identification were performed using Proteome Discoverer 2.5 software (Thermo Scientific) with a built-in Sequest HT search engine using a FASTA file containing 85 454 protein sequences of pigeons (TaxID 8930) downloaded from UniProtKB on 20.07.2022. Additionally, the data was validated against a database of typical contaminants. The crucial search criteria: precursor mass tolerance: 10 ppm; fragment mass tolerance: 0.05 Da; min. peptide length: 6; spectrum matching by b and y ions; fixed modifications: carbamidomethylation (C); variable modifications: oxidation (M), acetylation (protein N-term), Met-loss (protein N-term), Met-loss + acetylation (protein N-term). Extraction, purification and mass spectrometry analysis of glycosphingolipids Isolation, fractionation and mass spectrometry of glycolipids and orcinol staining were performed using standard procedures applied previously to human red blood cells [39] and 2102Ep cell lines [35]. Isolation, purification and identification of N- and O-glycans from pigeon RBC membranes Protein lysate was prepared using 3 mg of pigeon RBC membranes by adding 250 μl lysis buffer (250 mM Tris–HCl pH 8.0, 50 mM NaCl and 1X proteinase inhibitor cocktail) (Thermo Fisher Scientific, USA) and syringing 20 times. The lysate was centrifuged for 15 min at maximum speed (15 000 xg) at 40C. PNGase F (a kind gift from Prof. Mariusz Olczak) was used for the release of N-glycans according to Bereman et al. [40] with modifications and O-glycosidase (New England Biolabs, USA) for the release of O-glycans according to the manufacturer's protocol. Briefly, 50 μl of denaturing buffer (0.2% SDS, 2% β-mercaptoethanol) was mixed with 50 μl of protein lysate and incubated for 20 min at 980C. After cooling to room temperature, 27 μl of PBS pH 7.4, 15 μl of NP-40 and 8 μl of PNGase F were added. The reaction mixture was incubated overnight at 370C. N-glycans were purified as described before with modifications [41]. SPE C18 cartridge (Waters, USA) was washed with 10 ml of methanol and 5 ml of 0.5% acetic acid. Next, the sample was loaded on a cartridge, washed with methanol, eluted with 0.5% acetic acid, frozen, lyophilized and analyzed using LC–MS. O-glycans were eluted with 0.5% acetic acid and 80% acetonitrile solution, frozen, lyophilized and analyzed using mass spectrometry. MALDI-TOF analysis of O-glycans was conducted on an ultrafleXtreme MALDI TOF/TOF instrument (Bruker Daltonics, Germany). Lyophilized O-glycan samples were dissolved in water. DHB (2,5-dihydroxybenzoic acid) (Sigma, USA) was used as a matrix (100 mg/ml, ACN, 50 mM NaCl). Spectra were scanned in the range of 700–3500 m/z in the reflectron-positive mode. External calibration was applied using the peptide calibration standard (Bruker Daltonics, Germany). For identification, GlycoWorkbench 2.1 software and Glyco Mass Calculator (https://www.nist.gov/static/glyco-mass-calc/) were used. Fluorescent labeling of N-glycans N-glycans released by PNGase F were desalted by solid phase extraction (SPE) on Supelclean™ ENVI-Carb™ SPE columns (Sigma-Aldrich) according to Packer et al. [42]. Eluted glycans were dried down on SpeedVac, labeled and purified according to the previously described procedure [43]. Briefly, lyophilized N-glycans were incubated for 3 h at 65 °C with the labeling solution consisting of 2-aminobenzamide (2-AB; 60 mg/ml, Sigma-Aldrich) and sodium cyanoborohydride (60 mg/ml, Fluka) in acetic acid: DMSO (3:5). After cooling to RT, the excess dye was removed on Spe-ed Amide-2 SPE columns (Applied Separations, USA) and the glycans were eluted with ultrapure water and dried down by lyophilization. UPLC-MS analysis of RBC-derived N-glycans The 2-AB-labeled N-glycans were analyzed by ultra-performance liquid chromatography-mass spectrometry (UPLC-MS) on the Aquity I-Class Plus UPLC system coupled with the Vion® IMS-QTOF high-resolution mass spectrometer via an ESI source (Waters Corporation). N-glycans were dissolved in AcN: H2O (50:50) and separated on ACQUITY UPLC Glycan BEH Amide Column, 130 Å, 1.7 µm, 2.1 mM X 150 mM (Waters) at 60℃. The following gradient conditions were used: solvent A was 50 mM ammonium formate pH 4.4, solvent B was 100% acetonitrile (Chemsolv), time = 0 min (t = 0.0), 25% A, 0.4 ml/min; t = 35.0, 46% A, 0.4 ml/min; t = 36.0, 100% A, 0.2 ml/min; t = 39.5, 100% A, 0.2 ml/min; t = 43.5, 25% A, 0.2 ml/min; t = 47.6, 25% A, 0.4 ml/min; t = 55.0, 25% A, 0.4 ml/min. ESI source was tuned as follows: capillary voltage: 3.0 kV, source temperature: 120℃, desolvation temperature 350℃, desolvation gas flow: 800 L/h. For data collection, the fluorescence detector was set at Exλ 330 nm and Emλ 420 nm, and the mass spectrometer was operated in positive ion TOF mode. The ions from m/z 600 to m/z 2000 were registered with automatic real-time mass correction (Lock Mass reference standard, Waters). The UPLC runs were externally calibrated with a 2-AB-labeled glucose homopolymer standard (Waters). The chromatograms were integrated automatically with manual correction into individual peaks to which glycans were assigned based on glucose unit (GU) values and exact mass. The chromatographic and mass data were analyzed using the Waters UNIFI scientific information system with the integrated Waters Glycan GU Scientific Library. The GlycoWorkbench [44] was used to support glycan identification. Statistical analysis In total, 52 birds representing 10 species were evaluated (listed in Suppl. Table 1). To clarify the sampling rationale, species were included based on (a) their availability during routine diagnostic procedures, (b) their relevance to human-bird interactions (e.g., urban birds such as pigeons and crows, and pet parrots commonly seen in veterinary practice), and (c) the aim to include phylogenetically diverse representatives of Neoaves while remaining within ethical and logistical constraints. This approach ensured ethically compliant, opportunistic sampling while capturing biologically meaningful interspecies variability. All analyses were performed using GraphPad Prism [version 8.0.0 for Windows, GraphPad Software, San Diego, California, USA, www.graphpad.com] and R [45]. Interspecies and multi-species comparisons were conducted using non-parametric Kruskal–Wallis tests. When significant, post hoc pairwise comparisons were performed using Dunn’s test GraphPad Prism or in R with Benjamini–Hochberg correction (Benjamini–Hochberg false discovery rate (FDR) procedure). For each pairwise species comparison, Hodges–Lehmann estimators of the median difference and their 95% confidence intervals were calculated in R using the FSA package). Additional Hodges–Lehmann estimates and confidence intervals were obtained for descriptive purposes using non-parametric procedures implemented in the coin package. Exact adjusted p-values and confidence intervals are provided in Supplementary Table 4. Results Galα1 → 4Gal-terminated antigens are present on RBCs in Neoaves but not in Galliformes or Anseriformes birds Previous data about antigens with terminal Galα1 → 4Gal moieties in birds were derived only from analyses of liver, leucocytes and egg white glycoproteins [46, 47]. However, avian red blood cells were never evaluated. The importance of glycoconjugate RBC antigens in immunohematology and pathogen adhesion and distribution in host organisms warrants investigation. We examined RBCs of 55 birds belonging to 12 avian species by quantitative flow cytofluorometry, using a mouse anti-P1 monoclonal antibody (specificity: P1 and Gb3) (Suppl. Table 1; Fig. 1A). We found that only birds belonging to the Neoaves parvclass express Galα1 → 4Gal structures, with marked interspecies and interindividual variability (Fig. 1B). No binding of anti-P1 antibodies to RBCs of species belonging to the Galliformes and Anseriformes orders was observed.Fig. 1Agglutination-based immunoassays and distribution of P1/Pk antigens among various avian species. A: Cytofluorometric quantitative analysis of anti-P1/Pk antibody binding (clone 650) to erythrocytes of various bird species. B: Comparison of the number of P1, Pk antigens on erythrocytes of two breeds of pigeons. C: Binding of Shiga Toxin subunit 1B (Stx1B) to avian erythrocytes. E: Levels of anti-P1/Pk antigens antibodies in avian sera. Scientific names of examined species are summarized in Suppl Table 2. F: Agglutination of parrots and pigeon' erythrocytes with sera of birds of different species. G: Agglutination tests for Scarlet Macaw's erythrocytes with sera of birds of selected species. H: Agglutination inhibition test on Scarlet Macaw's erythrocytes with pigeon egg's ovomucoid bearing P1-glycotopes for selected avian sera. I: Agglutination of erythrocytes of individuals with different blood types by sera of birds of different species. P1PK genotypes (P1P1, P2P2) and ABO phenotypes (A, B, and O) are given. J: Analysis of human glycophorin A binding to pigeon serum in ELISA tests Agglutination-based immunoassays and distribution of P1/Pk antigens among various avian species. A: Cytofluorometric quantitative analysis of anti-P1/Pk antibody binding (clone 650) to erythrocytes of various bird species. B: Comparison of the number of P1, Pk antigens on erythrocytes of two breeds of pigeons. C: Binding of Shiga Toxin subunit 1B (Stx1B) to avian erythrocytes. E: Levels of anti-P1/Pk antigens antibodies in avian sera. Scientific names of examined species are summarized in Suppl Table 2. F: Agglutination of parrots and pigeon' erythrocytes with sera of birds of different species. G: Agglutination tests for Scarlet Macaw's erythrocytes with sera of birds of selected species. H: Agglutination inhibition test on Scarlet Macaw's erythrocytes with pigeon egg's ovomucoid bearing P1-glycotopes for selected avian sera. I: Agglutination of erythrocytes of individuals with different blood types by sera of birds of different species. P1PK genotypes (P1P1, P2P2) and ABO phenotypes (A, B, and O) are given. J: Analysis of human glycophorin A binding to pigeon serum in ELISA tests Next, we checked for the presence of anti-P1/Pk antibodies in the sera of wild and pet birds. The level of antibodies detected was relatively low and was found in only half of the tested species. Nevertheless, a significant variation in anti-P1 antibody levels was observed across the analyzed sera. A notable interspecies distribution was evident between pigeons and parrots, the most represented species in the study, due to their relevance to human activities, particularly in breeding and exhibition contexts. The most striking finding was the significant difference observed among the representatives of the four pigeon breeds examined, which included two domesticated and two wild-living breeds (Fig. 1E). Antigen levels differed significantly among species (Kruskal–Wallis H = 61.23, p < 0.0001. Post-hoc Dunn tests identified several significant pairwise differences in antigen levels between species (see Supplementary Table 4 for exact adjusted p-values and 95% confidence intervals for Hodges-Lehmann differences). We carried out a series of interspecies agglutination and agglutination inhibition tests using avian RBCs (Common Cuckoo, Racing Pigeon, Black-And-Yellow Macaw, Catalina Macaw, And Scarlet Macaw) and avian sera (21 birds of different species, Fig. 1F). RBCs of birds kept as pets, being in close contact with humans, agglutinated the highest number of about 16 of 21 sera (All names of tested avian species are listed in Suppl. Table 2). The strongest reactions occurred for two parrot species, Catalina Macaw and Scarlet Macaw, particularly when compared to sera of representatives of other bird orders Accipitriformes (Montagu's Harrier), Piciformes (Great Spotted Woodpecker), Columbiformes (Racing Pigeon, Rock Pigeon, Woodpigeon), Apodiformes (CommonSwift), Accipitriformes (Red Kite), Psittaciformes (White Cockatoo), and Passeriformes (Hooded Crow) (Fig. 1F). Our study aimed to determine if agglutination was due to the presence of anti-P1 alloantibodies or if other factors were involved in this process. To verify this hypothesis, a pigeon's egg ovomucoid solution containing the P1 antigen was used as an agglutination inhibitor reagent. Sera of four different orders' representatives exhibited agglutination inhibition upon utilizing an inhibition reagent. Those sera were tested extensively with scarlet macaw RBCs, and the results showed that the anti-P1 alloantibodies level for sera of representatives of Columbiformes (racing pigeon) and Passeriformes (hooded crows) is surprisingly elevated and can be overcome only for the highest concentrations of agglutination inhibitor. In the remaining two orders, agglutination inhibition is observed even for the lowest concentration, indicating a low level of anti-P1 alloantibodies on scarlet macaw RBCs (Fig. 1G – H). To find out if the Galα1 → 4Gal glycotopes recognized by the mouse anti-P1 antibody on RBCs were carried by glycosphingolipids or glycoproteins, the neutral glycosphingolipid fractions from avian RBCs were isolated and analyzed by high-performance thin-layer chromatography (HPTLC) (Fig. 2A). Orcinol staining revealed that GlcCer and LacCer were the major neutral glycosphingolipids in C. livia (Neoaves). In addition, the neutral GSL fraction from fancy pigeon contained Gb3Cer and a GSL with mobility similar to the human P1 blood group antigen (Fig. 2A). In HPTLC immuno-overlays, human anti-P1 antibody recognized an extra GSL. The double P1 bands recognized by the human anti-P1 antibody most likely represent multiple P1 isoforms with different fatty acids [48]. No GSLs reactive with anti-NOR antibodies were detected (Fig. 2B). In RBCs from Galloansers species (chicken, goose) and racing pigeons, we did not detect any binding of anti-P1 antibodies.Fig. 2Analysis of glycosphingolipids, proteins, and glycans from erythrocytes of birds. A-B: HPTLC analysis of neutral glycosphingolipid fractions isolated from erythrocytes of modern birds and terrestrial poultry. (A) Orcinol staining and anti-P1 binding detection (clone 650 recognizing P1 and Pk antigens and clone P3NIL100 recognizing P1 antigen) and (B) binding of Shiga toxin 1 subunit B (Stx1B) and anti-NOR antibody (nor118). C: Binding of anti-P1 antibodies (clone 650 and P3NIL100) and Shiga toxin 1 subunit B (Stx1B). E: mass spectrum of GSLs isolated from fancy pigeon erythrocytes; F: mass spectrum of GSLs isolated from racing pigeon erythrocytes; G: mass spectrum of GSLs isolated from chicken erythrocyte membranes. H: LC-MS-based identification of N-glycans isolated from racing pigeon (C. livia) erythrocyte membranes. In red circles, N-glycans with terminal Gal-Gal moieties and B: binding of Shiga toxin 1 subunit B (Stx1B) and anti-NOR antibody (nor118) Analysis of glycosphingolipids, proteins, and glycans from erythrocytes of birds. A-B: HPTLC analysis of neutral glycosphingolipid fractions isolated from erythrocytes of modern birds and terrestrial poultry. (A) Orcinol staining and anti-P1 binding detection (clone 650 recognizing P1 and Pk antigens and clone P3NIL100 recognizing P1 antigen) and (B) binding of Shiga toxin 1 subunit B (Stx1B) and anti-NOR antibody (nor118). C: Binding of anti-P1 antibodies (clone 650 and P3NIL100) and Shiga toxin 1 subunit B (Stx1B). E: mass spectrum of GSLs isolated from fancy pigeon erythrocytes; F: mass spectrum of GSLs isolated from racing pigeon erythrocytes; G: mass spectrum of GSLs isolated from chicken erythrocyte membranes. H: LC-MS-based identification of N-glycans isolated from racing pigeon (C. livia) erythrocyte membranes. In red circles, N-glycans with terminal Gal-Gal moieties and B: binding of Shiga toxin 1 subunit B (Stx1B) and anti-NOR antibody (nor118) The MALDI-TOF mass spectrometry revealed that GSLs of chicken, goose, fancy and racing pigeons contained several clusters of monosodiated molecular adduct ions [M + Na] + corresponding to glucosylceramide (GlcCer) (Fig. 2E – G). The mass differences between the ions (mostly ~ 28 Da, which is the molecular mass of two methylene groups) correspond to isoforms with ceramides containing acyl groups of different lengths (e.g., 16:0, 18:0, 24:0). In the GSL sample from chicken (Fig. 2G), we identified only ions corresponding to GlcCer, from 725.6 m/z to 851.6 m/z [16:0 to 24:1 acyl group, also with unusual chain length (23:0) and/or hydroxylated, respectively]. In the fancy pigeon RBCs (Fig. 2E), we found ions for ceramide (mostly hydroxylated fatty chains), GlcCer (the largest observed group), and ions at m/z 957.6, m/z 985.7, m/z 1013.4, m/z 1199.8, 1343.9 and m/z 1365.9 corresponding to LacCer (20:0 h, 22:0 h, 24:0 h), nLc4Cer (15:0), and fucosylated nLc4Cer (15:0, 16:3) respectively. Thus, GlcCer was the predominant component of total neutral glycosphingolipids in avian RBCs (Fig. 2F – G). Structures recognized by the mouse anti-P1 antibody (bands representing proteins with MW 80 kDa and higher) were found to be present only in lysates from RBCs of birds belonging to Neoaves parvclass, and we saw a similar pattern in Western blotting for both analyzed samples. In contrast, Western blotting analysis of RBC lysates from birds belonging to Galloansers parvclass and Paleognathae infraclass with the human anti-P1 antibody revealed only nonspecific binding (Fig. 2C). To evaluate the glycan structures, we performed the LC–MS N-glycome analysis. Based on GU values and exact mass confirmation, 30 distinct glycan structures and compositions were identified (Fig. 2H, Suppl. Data 3). When no mass data was available, the chromatographic peaks were left unassigned, which was the case with weak signals. Structural isomers were not indicated in cases where GU data were insufficient. The predominant peak within the fluorescence chromatogram was identified as M9Glc1 (GU 10.18). The other significant signals were assigned as oligomannosidic N-glycans from M3 to M9 at the GUs: 4.16, 5.15, 6.11, 6.97, 7.83, 8.73 and 9.45, respectively. The other identified species were hybrid glycans and complex di-antennary structures, with the predominant signal from biantennary, disialylated glycan (A2G(4)2S(3,6)2) at GU 9.29. A total of 6 glycans were core-fucosylated, 9 were sialylated, and 6 were modified with the bisecting GlcNAc residue. Additionally, four N-glycans with possible terminal, alpha-bound Gal residues were identified (A2G(4)1Ga1, GU 7.07; F(6)M4A1G(4)1Ga1, GU 8.00; A2G(4)2Ga1S1, GU 8.89; F(6)A2G(4)2Ga2, GU 9.01). MALDI TOF O-glycome analyses revealed over 50 structures (summarized in Suppl. Data 4). A total of 34 O-glycans were core-fucosylated, 19 were sialylated, and 7 potentially had terminal Gal-Gal residues. Core-1 and 3 were the most abundant among the identified structures. Antibodies against the human GPA and P1 antigens are present in the blood sera of wild birds Next, we asked whether antibodies recognizing the human P1PK blood group antigens may be present in birds. In humans, anti-P1PK antibodies are common in P2 individuals. To evaluate the presence of such antibodies in birds, we performed agglutination tests using avian sera and human RBCs. Avian sera in various dilutions were incubated with human RBCs of blood groups P1, P2, as well as A, B, and O, in multiple combinations (Fig. 1I). P1 (P1P1/A), P2 (P2P2/A), and P2 (P2P2/O) RBCs were agglutinated the most. Sera from wild birds and Princess Parrots (Polytelis alexandrae) did not agglutinate human RBCs, while sera from Blue-and-Yellow Macaw (Ara ararauna) and Harlequin Macaw (Ara harlequin) (parrots) agglutinated P1O, P2A and P2A RBCs treated with papain and, to a lesser extent, other phenotypes. The P1O RBCs were agglutinated the least (Fig. 1I). To determine whether the avian antibodies recognized the P1 or Pk antigen, we performed an agglutination inhibition test using pigeon ovomucoid as an inhibitor because it contains glycoproteins with P1 glycotopes (Suppl. Table 3) [46, 49]. No agglutination inhibition was observed, suggesting these antibodies recognize antigens unrelated to the P1PK group system. To answer the question of whether these antibodies recognize human AB (H) group antigens, we performed an agglutination inhibition test using human glycophorin A isolated from type A RBCs because N- and O-glycans of that protein contain the A antigen [50]. We found that glycophorin A inhibited agglutination, suggesting that the agglutinating antibodies in the avian sera could recognize the A antigen (Suppl. Table 3). Glycophorin A from a type O individual also inhibited agglutination. Inhibition was only observed for RBCs treated with papain, a nonspecific protease, indicating that the pigeon serum contained antibodies that recognized the carbohydrate chains of the human A/B/H antigens located on glycophorin, rather than the protein moiety of glycophorin. Using ELISA, we found that the antibodies from the rock pigeon serum bound to human glycophorin A isolated from human type A RBCs (Fig. 1J). Thus, we propose that rock pigeons and pet parrots (Blue-and-yellow macaw, Cross-breed macaw, Scarlet macaw) make antibodies recognizing the human ABO but not P1PK blood group antigens. Rab11FIP5 as a potential receptor for Shiga toxins? Since the antigens of the human P1PK blood group system are receptors for Shiga toxins, we evaluated the binding of Shiga toxin subunit 1B to RBCs from different avian species (Fig. 1C). Stx1B bound to RBCs from Neoaves parvclass species, with the strongest binding detected for peregrine falcon (Falco peregrinus). No significant differences were found between other species. To determine if Shiga toxin recognizes GSLs from avian RBCs, we performed HPTLC immunooverlay assays of neutral glycosphingolipids isolated from pigeon, chicken, and goose RBCs. Stx1B is bound only to fancy pigeon GSLs (Fig. 2B). In Western blotting analysis prepared on 2D-electrophoresis gel, binding of Stx to a single band with a molecular weight of approximately 80 kDa was observed. Binding to proteins with a molecular weight lower than 34 kDa was found to be nonspecific. In our opinion, Stx1 recognized a single protein containing N-glycans with Galα1 → 4Gal terminal structures, which is a putative receptor for Shiga toxins [8]. To identify the protein recognized by Stx1B in pigeon RBC membranes in Western blotting, we used LC–MS analyses of a corresponding 2DE protein spot on a two-dimensional gel band. The analysis showed that the recognized protein was Rab11-family interacting protein 5 (Rab11FIP5) (Suppl. data 1, Suppl. data 2). This 66 kDa protein, with two predicted N-glycosylation sites and multiple phosphorylation sites (also predicted using in silico analysis [48, 49]), is a well-described endosomal trafficking regulator that forms complexes with Rab11 on endosomal membranes, directing cargo for degradation in lysosomes [51, 52]. As a member of the RAB11FIP family, RAB11FIP5 specifically interacts with the small GTPase RAB11A, which is a key regulator of vesicle trafficking from the recycling endosomes to the plasma membrane [53, 53, 54]. This interaction is crucial for the proper localization and function of RAB11A, influencing various cellular processes, including receptor recycling, polarized protein transport, and cytokinesis. Additional research is required to verify the specificity of Stx1B binding across a broader spectrum of avian species and to elucidate the functional implications of Rab11FIP5 recognition by Shiga toxin in avian cells. This novel idea of an additional non-human Shiga toxin non-glycosphingolipid receptor necessitates broadening the view of expertise in 'avian omics,' which is necessary to gain a detailed understanding of the phenomenon discussed. Discussion Avian blood group systems have been understudied despite their relevance for human and pet health. While human Gb3/CD77 synthase has been the subject of numerous studies by our group and others [6–8, 39, 55, 56], the only well-described homolog from a non-human species is the pigeon Gb3/CD77 synthase from C. livia. Unlike human Gb3/CD77 synthase, the pigeon enzyme primarily transfers galactose to glycoprotein acceptors, exhibiting low activity with GSL acceptors [57, 58]. Galα1 → 4Gal structures were found in modern birds (Neoaves) but not in Galliformes/ratites (Galloansers and Paleognathae) [47, 59]. All data on Galα1 → 4Gal in birds published to date have come from analyses of glycoproteins isolated from egg white, liver, or lymphocytes [60]. Western blotting experiments with Griffonia simplicifolia-I lectin (specific for terminal αGal/GalNAc) and anti-P1 monoclonal antibody (specific for Galα1 → 4Galβ1 → 4GlcNAcβ1) showed that birds belonging to Struthioniformes, Tinamiformes, Cracidae, Galliformes and Anseriformes do not express Galα1 → 4Gal moiety on glycoproteins. This finding suggests that Galα1 → 4Gal expression went under evolutionary pressure in modern birds, and only Neoaves can produce Galα1 → 4Gal-capped glycans [47]. The P1PK blood group system is one of several immunological factors influencing blood compatibility in birds, with particular relevance to transfusion medicine. Accurate identification of these antigens is essential to avoid immune reactions during blood transfusions in birds. Understanding species-specific P1PK profiles can enhance transfusion safety, and improve clinical outcomes in critical care. In contrast to in-depth studies of the P1PK blood group system in humans, avian RBCs have never been tested for expression of the P1PK antigens. In this study, we evaluated RBCs from over 50 individuals belonging to 12 bird species using flow cytofluorometry, Western blotting, HPTLC, and mass spectrometry. RBCs from urban birds (pigeons, crows) or pets (parrots) showed the highest expression (Fig. 1A – B). We found the lowest expression in birds of prey (falcon) and game species birds (woodcock). Our study found distinct differences in P1PK antigen expression between bird species from different habitats. Urban birds, such as pigeons or parrots kept as pets, showed higher levels of these antigens than rural birds. This could be an evolutionary adaptation to the challenges of city environments, which expose them to more pathogens and environmental stresses. Rural birds might use different strategies to cope with their habitat. Comparing these results with earlier research helps us better understand how habitat impacts P1PK antigen expression and what this means for bird health and disease resistance [61]. Interestingly, Stx1B showed the highest binding to falcon RBCs, with relatively low binding to RBCs from other species (Fig. 1C). Neither anti-P1 antibodies nor Stx1B recognized RBCs from ratites (terrestrial poultry, pheasant). In pigeons, we found that some of the detected P1PK antigens are GSLs Pk and P1. These results were confirmed by mass spectrometry. Our findings corroborate the previous data showing that only Neoaves can express Galα1 → 4Gal moieties. Since birds belonging to Neoaves express large amounts of the P1PK blood group antigens, we asked whether birds develop alloantibodies recognizing these antigens, similarly to humans. To date, avian blood group differentiation has been studied primarily in the chicken (Gallus domesticus) [18, 62]. We found anti-P1 or anti-Pk antibodies in a few of the avian sera that we tested, but on a low level (Fig. 1E). Moreover, sera from birds living close to humans contained antibodies recognizing the A, B, and H antigens from human RBCs (Fig. 1I). This raises the question of how those birds develop antibodies against human blood group antigens. Most likely, immunization occurs through exposure to human microbiota, which is known to play a crucial role in the development of natural anti-A/B/H antibodies [63–65]. As no ABO homologs have been identified in birds [25], the antibodies that we detected are unlikely to be alloantibodies. The interactions between birds and humans, especially concerning P1PK antigens, may have more health importance than previously thought. Since some bird species are known to carry zoonotic pathogens, understanding how P1PK works in these interactions could help identify potential health risks. People who regularly contact birds, such as veterinarians, breeders, or bird keepers, may unknowingly be exposed to antigens that can cause immune or allergic reactions. The patterns of interspecies agglutination with avian red blood cells and sera from various bird species were investigated to determine if anti-P1 alloantibodies or other factors largely influenced agglutination. Obtained results revealed that bird species kept as popular pets, specifically Catalina macaw and scarlet macaw, demonstrated the most substantial level of agglutination when exposed to sera from different bird species (Fig. 1F). Our findings validate earlier reports of avian species possessing naturally high titers of various alloantibodies due to feasible evolutionary immunological mechanisms or exposure to foreign antigens in the environment [66–69]. This indicates variation in immune evolution and antigen exposure across avian taxa. These ecological trends should be interpreted cautiously, as factors such as diet composition, microbiota structure, and environmental microbial exposure are potential confounders that may also contribute to variation in glycan expression between urban and rural bird populations [70, 71]. The increased reactivity observed in the red blood cells of Catalina Macaws and Scarlet Macaws in our preliminary screening needs further investigation. These species exhibited significantly greater agglutination than other examined species. The primary hypothesis assumes that specific P1-like glycoprotein or glycolipid structures on the surface of their RBCs may contribute to observed antigenicity. Furthermore, the potential for cross-reactivity should be assessed across a broader range of species to determine whether these macaws possess distinctive RBC surface markers that enhance agglutination. Since little is known about avian immunohematology, blood transfusions in birds are not a common clinical practice. A high rate of serologic incompatibility has been observed in various avian species, particularly in parrots. Most medical knowledge about birds originated from representatives of Psittaciformes, the most common pet birds. Most avian patients are treated for anemia. Interestingly, avian species tend to tolerate hematological diseases better than mammals. The ability of birds to endure massive blood losses is one of the reasons why blood transfusions are typically limited to life-threatening cases, where blood loss reaches 30% of the total blood volume. Due to the lack of a blood group system classification in birds, the standard blood transfusion protocol recommends donors from the same species, when available, or the phylogenetically closest species. Data on RBC survival after heterologous blood transfusion or transfusion reaction symptoms in birds are limited, suggesting that homologous transfusion may be preferred over heterologous transfusion in pet birds [24, 66, 72]. Occurrence of agglutination of the sera of birds belonging to other orders is an expected incident previously described (Degernes et al., 1999; Hohenhaus, 2004; Martinho, 2009; Shaw et al., 2009), but herein we presented another proof for intraspecies blood groups discrepancies which play a crucial role in avian transfusion compatibility and immune response. These findings highlight the complexity of avian blood group interactions and emphasize the need for further research to address the knowledge gap regarding the underlying factors contributing to this phenomenon. They also explore their potential applications in simplifying and increasing the efficiency of pretransfusion protocols regarding birds in veterinary medicine. Bird keepers and bird health professionals should understand the impact of P1PK antigen expression on their birds' health. Regular veterinary visits, including blood tests to determine P1PK compatibility, may help to prevent complications, particularly during transfusions [73]. Shiga-producing enterohemorrhagic E. coli (STEC) strains periodically cause potentially deadly outbreaks of hemorrhagic colitis and hemolytic uremic syndrome (HUS). The main receptor for toxins, Gb3, is found in the human vascular endothelium, which plays a crucial role in the sequelae of STEC infection. Upon entry into the bloodstream, the pentameric subunits of Shiga toxins bind to the globo series GSL, showing a strong preference for the disaccharide Galα1 → 4Galβ1. Studies to identify the non-glycosphingolipid receptor for Stx on human neutrophils have shown that it is the TLR4 molecule [74]. The interaction of Stx with TLR4 suggests a mechanism of protein–protein interaction between the A subunit of Stx and TLR4 that may act independently of the protein–sugar interaction between the B pentamer of Stx and Gb3Cer [75, 76]. Gb3 is considered to be the primary receptor for Stx holotoxins. The strength of Stx binding to the host cell is also influenced by the composition of the lipid bilayer [77, 78]. Due to its binding to different microdomains in the cell membrane, Stx can be transported to the cytosol via various routes [79]. Once bound to the cell surface, toxins can be internalized through several mechanisms. Once Gb3 is recognized, Shiga toxin can enter the cell through endocytosis and be transported to the cytosol. The ability to inactivate the 60S subunit of the ribosome renders Stx a potent inhibitor of protein synthesis [80]. Enzymatic inactivation of eukaryotic ribosomes results in irreversible inhibition of cellular protein biosynthesis. This is especially true for those ribosomes that are bound to mRNA and are biochemically active at the time Stx cleaves the glycosidic bond. Cytotoxic effects include the inhibition of translation [81] and the activation of a ribotoxin stress response, which induces damage to the ribosome through the A1 subunit. Ribotoxic stress and ER-related stress are thought to play a more significant role in Stx cytotoxicity than inhibition of protein synthesis [82]. It has been demonstrated that Shiga toxins, in addition to ribosome inactivation, can alter the phenotype of endothelial cells, influencing the transcriptional regulation of certain genes. The molecular mechanisms responsible for these changes remain unknown [83, 84]. This inhibition of protein synthesis triggers a cascade of cellular processes that ultimately alter the phenotype of endothelial cells. Phenotypic modifications induced by Shiga toxin in endothelial cells are various and intricate, ranging from alterations in cell morphology and adhesion molecule expression to the production of inflammatory mediators and induction of apoptosis [85, 86]. Shiga toxin's interference with endothelial cell function is central to the pathogenesis of hemolytic-uremic syndrome, a devastating condition characterized by microangiopathic hemolytic anemia secondary to E. coli infection, thrombocytopenia, and acute kidney injury. Shiga toxin's actions extend beyond increased cytotoxicity, as it also influences endothelial cell permeability, resulting in edema and inflammation within target organs [87, 88]. Endothelial cells, the inner lining of blood vessels, play a crucial role in regulating vascular homeostasis, blood flow, and mediating inflammatory responses. When exposed to Shiga toxin, these cells undergo a series of phenotypic changes that compromise their normal functions [89]. These changes include increased expression of adhesion molecules such as E-selectin, ICAM-1, and VCAM-1, which enhance leukocyte recruitment and adhesion to the endothelial cell, thereby increasing inflammation. Furthermore, Shiga toxin can initiate the release of proinflammatory cytokines and chemokines, such as interleukin-1β, interleukin-6, and TNF-α, which further intensify the inflammatory response and also contribute to endothelial cell activation and dysfunction [90–92]. The toxin also initiates morphological alterations in endothelial cells, including rounding of cells, detachment from the extracellular matrix, and destruction of intercellular junctions. These morphological alterations increase endothelial permeability, leading to the leakage of plasma proteins and fluid into surrounding tissues, which results in edema. Shiga toxin action involves a complex interaction of signaling cascades that regulate the cytoskeleton, microtubule reassembly, and dynein-dependent transport. Rho GTPases, responsible for regulating endocytosis and retrograde motility of Shiga toxin, play key roles in these processes. Shiga toxin movement via the secretory pathway is dependent on the cytoskeleton [93]. We sought to identify the glycoprotein receptor for Stx1 in Neoaves and found a membrane glycoprotein that we propose to be the avian Rab11-family interacting protein 5 (Rab11FIP5). Rab11FIP5 is an effector protein and partner of Rab11 in a complex responsible for cargo transport in recycling endosomes, mediating degradation in lysosomes or exocytosis (in neuroendocrine cells), cytoskeleton rearrangement and iron uptake [94, 95]. FIPs are a family of proteins that form complexes with Rab11. These complexes are essential for the proper function of Rab11, as they link it to molecular motors and play a role in recycling endosome transport [54, 96]. Rab11, a GTPase, is a protein belonging to the Ras-like small GTP-binding protein superfamily and is best known as a major regulator of membrane trafficking in cells (e.g., phagocytosis, cytokinesis) [53]. RAB11FIP5 contains several functional domains, including an N-terminal domain involved in RAB11 binding, a central coiled-coil domain that mediates protein–protein interactions, and a C-terminal domain implicated in cargo recognition. The intricate interplay between these domains enables RAB11FIP5 to act as a scaffold protein, coordinating the assembly of protein complexes that facilitate the efficient trafficking of specific cargo molecules through the endocytic recycling system. [97, 98]. It is beyond question that the exact mechanisms by which RAB11FIP5 modulates Shiga toxin trafficking are complex and multi-layered, involving interactions with other trafficking proteins and regulatory molecules. One of our proposed hypotheses suggests that it is plausible that RAB11FIP5 facilitates the recycling of the Shiga toxin receptor, redirecting it back to the cell surface and thereby prolonging the toxin's availability for binding and internalization. Furthermore, RAB11FIP5 might participate in forming specialized transport carriers that mediate the trafficking of Shiga toxin between different intracellular compartments. This process could involve the recruitment of other proteins, such as motor and adaptor proteins, which facilitate the movement of toxin-containing vesicles along microtubule tracks. Birds are reservoirs of STEC, but surprisingly, they do not show any clinical infection symptoms regardless of Shiga toxin receptor expression. Several mechanisms may determine avian insusceptibility to Shiga toxins [99, 100]. Low levels of the main receptor Gb3 and/or its poor availability for toxin molecules may be one of the resistance factors. Additionally, the membrane topology of avian cells may hinder the access of toxins to receptors. Many reports have suggested that membrane topology, glycosylation of its components, or cholesterol content may affect the access and nature of the interactions between the toxins and receptors [76, 80, 101–104]. Additionally, most avian GSLs analyzed in this study have short-chain fatty acids in Gb3 which may contribute to resistance, while P1 glycotopes could act as decoy receptors, binding toxin without triggering cytotoxicity [105, 106]. Our preliminary findings suggest a novel mechanism of Stx nullification in avian cells, potentially involving a Golgi-bypassing uptake pathway and Rab11FIP5-mediated rerouting. The involvement of intracellular trafficking proteins such as Rab11FIP5 is intriguing; however, this remains hypothetical and requires targeted experimental validation. We therefore present this as a potential direction for future work rather than a confirmed mechanism. Future research should investigate the genetic factors regulating P1PK antigen expression in birds. Exploring how environmental conditions affect these genes might clarify variation among species. Examining P1PK’s role in defense against pathogens, including STEC, may reveal new insights into avian immunity and potential therapeutic strategies. Despite progress in understanding how Stxs recognize, enter, and kill human cells, there is an unmet clinical need for treatment options. Avian resistance mechanisms could inspire novel therapeutic approaches for STEC infections, continuing public health concerns. Study limitations While our findings shed light on the presence and potential role of P1PK antigens in bird species, several limitations need to be acknowledged. Although the study included a relatively diverse range of species, the number of individuals per species was limited, which may limit the broad applicability of the results. The absence of functional assays in bird cell lines also limits our ability to determine the intracellular fate and cytotoxic effects of Shiga toxins in birds with certainty. Since the study focused only on red blood cells, other cell types, such as endothelial or epithelial cells, possibly involved in host–pathogen interactions with P1PK-like antigens, will be explored in future research. The identification of Rab11FIP5 as a potential Shiga toxin–binding protein through LC–MS/MS requires further validation for biological relevance. A key limitation is the incomplete annotation of bird proteomes, particularly for non-model species like pigeons (Columba livia). Many pigeon proteins are predicted based on sequence similarity to better-studied organisms, such as chicken or zebra finch, without experimental verification or detailed functional annotation. This limits the depth of proteomic analysis and increases the risk of misannotation. Additionally, the functional role of Rab11FIP5 as a possible non-GSL receptor for Shiga toxin and its role in endosomal trafficking in bird cells are beyond the scope of this work. Future research should focus on targeted biochemical and cellular experiments, preferably using bird-specific cell culture models. Supplementary Information Supplementary Material 1: Suppl. data 1. LC-MS_2DE Samples identification lists (raw data). Suppl. data 2. Base Peak and Total Ion Chromatograms (BPC & TIC) of the analyzed 2DE samples. Suppl. Data 4. List of predicted O-glycan structures identified from pigeon erythrocytes. Supp. Table 1. Binding of P1/Pk antibody (clone 650) to avian erythrocytes. Names of Galliformes representatives are marked in red.Suppl. Table 2. Common and Scientific names of avian species used in experiments.Suppl. Table 3. Agglutination inhibition screening test of human erythrocytes by pigeon serum. Suppl. Fig.1. Representative flow-cytofluorometry gating plot (corresponding to Fig. 1A-C). Supplementary Material 2. Supplementary Material 1: Suppl. data 1. LC-MS_2DE Samples identification lists (raw data). Suppl. data 2. Base Peak and Total Ion Chromatograms (BPC & TIC) of the analyzed 2DE samples. Suppl. Data 4. List of predicted O-glycan structures identified from pigeon erythrocytes. Supp. Table 1. Binding of P1/Pk antibody (clone 650) to avian erythrocytes. Names of Galliformes representatives are marked in red.Suppl. Table 2. Common and Scientific names of avian species used in experiments.Suppl. Table 3. Agglutination inhibition screening test of human erythrocytes by pigeon serum. Suppl. Fig.1. Representative flow-cytofluorometry gating plot (corresponding to Fig. 1A-C). Supplementary Material 2.