Pathological evidence of neurotropism and oculotropism in wild black-headed gulls naturally infected with H5N1 high pathogenicity avian influenza Authors: Manuela Crispo, Nicolas Gaide, Guillaume Croville, Mathilda Walch, Maxime Fusade-Boyer, Malorie Dirat, Cecile Caubet, Grégory Jouvion, Pascal Arné, Irene Valverde Dominguez, Loïc Palumbo, Quentin Bigot, Jean Luc Guerin, Julien Hirschinger, Guillaume Le Loc’h Published: 2026 ABSTRACT Clade 2.3.4.4b H5Nx high pathogenicity avian influenza viruses (HPAIVs) have caused extensive mortality in wild birds, particularly colonial seabirds, since 2020. Among them, black-headed gulls (Chroicocephalus ridibundus) are abundant, synanthropic birds widely distributed across Europe and northern Asia that exploit diverse habitats, raising questions about their role in AIVs ecology. Recent field observations have reported post-HPAI ocular abnormalities in some marine birds, suggesting a potential contribution of the eye to disease expression. However, the nature of these changes, viral tissue distribution and the underlying pathogenesis remain unclear. To address this, 13 black-headed gulls from a cohort of 49 confirmed HPAI cases in France underwent clinico-pathological assessment, immunohistochemistry and molecular analysis. Affected birds commonly exhibited emaciation and central neurological signs. Histopathology and immunostaining revealed marked neurotropism and consistent involvement of intraocular structures, including anterior uveitis and retinal degeneration. Viral antigen was detected in neuronal tissue, visceral and integumentary epithelial cells, with no evidence of endotheliotropism. This pattern resembles that reported in domestic ducks and may influence viral diffusion, persistence, intra- and interspecific transmissions. Our results suggest the eye is an underestimated target organ in HPAI infection, and provide new insights into the pathogenesis of HPAI-induced ocular abnormalities in seabirds.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-57262-2. Introduction Since the end of 2020, high pathogenicity avian influenza (HPAI) H5Nx viruses belonging to clade 2.3.4.4b have caused unprecedented epizootics in wild birds, with an expanded host range including wild mammals, companion animals and livestock1–4. Among the over 500 wild bird species affected, colonial seabirds have been particularly impacted. Heavy losses have been reported across the globe, particularly among northern gannets (Morus bassanus), Peruvian boobies (Sula variegata), guanay cormorants (Leucocarbo bougainvilliorum), great skuas (Stercorarius skua), sandwich terns (Thalasseus sandvicensis), Humboldt penguins (Spheniscus humboldti), and Peruvian pelicans (Pelecanus thagus), raising serious concerns about ecosystem disruption, and the long-term survival of endangered species5–11. Gull species, including black-headed gulls (Chroicocephalus ridibundus), were not spared11–15. In addition to acute mortality, field observations have recently reported ocular abnormalities in marine birds. Diffusely black or mottled irises (“black eyes”), in contrast to their original pale blue/grey color, and anisocoria were recently described in serologically AIV-positive northern gannets (Morus bassanus)16,17. Neurological signs are also frequently observed during the acute phase of disease, and neurotropism of H5N1 clade 2.3.4.4b viruses has been demonstrated in multiple avian species18. Overall, these findings suggest that both the central nervous system and the eye may be involved in HPAI pathogenesis and disease expression16; however, the mechanisms underlying ocular lesions, including viral dissemination, tissue tropism, and functional consequences, remain poorly understood. To date, similar ocular changes have not been documented in gulls or other wild bird species. Black-headed gulls are medium-sized, synanthropic birds, widely distributed across Europe and northern Asia and are recognized as important hosts in the ecology of avian influenza viruses, with documented susceptibility to both low and high pathogenicity strains, including gull-associated subtypes19–22. They breed in freshwater wetlands, lakes, rivers and coastal areas and can be found in Western Europe, including France, extending inland, to urban centres and agricultural fields23. In 2021, their population in France was estimated at 279,653 wintering individuals24. Black-headed gulls are opportunistic feeders, consuming fish, insects, aquatic invertebrates, and a variety of anthropogenic wastes25–27. They also engage in kleptoparasitism, stealing prey gathered by other birds28,29 and form large colonies that may attract other species, such as sandwich terns, seeking protection from predators22,30,31. During the winter 2022–2023, substantial mortality events in this species were recorded across Europe11–13,18,32,33. In France, dead birds were collected through the SAGIR network (“Surveiller pour AGIR”, monitoring for action), an event-based epidemiological surveillance system in charge of HPAI surveillance in wildlife. Concurrently, an increased number of clinically affected birds were received by many wildlife rehabilitation centers. Both mortality and morbidity were ultimately attributed to HPAI H5N1 clade 2.3.4.4b. infections following confirmation by molecular testing conducted at the National Reference Laboratory (ANSES’s Ploufragan-Plouzané-Niort Laboratory, France). This outbreak provided a well-defined and valuable opportunity to investigate tissue tropism and clinico-pathological presentation in naturally infected wild black-headed gulls, with a particular focus on central nervous system and ocular involvement. Materials and methods Ethics statement Live birds included in this study were free-living wild black-headed gulls (Chroicocephalus ridibundus) submitted for diagnostic investigation as part of wildlife disease surveillance in France. These animals were not part of any experimental trial. Individuals exhibited severe clinical signs and were humanely euthanized for animal welfare and biosecurity reasons by authorized veterinary personnel, in accordance with French regulations and relevant institutional and national guidelines. Considering the retrospective nature of this study, no institutional ethical approval was required. Clinical assessment Between December 2022 and February 2023, 49 black-headed gulls originating from northern France tested positive for HPAIV by molecular analysis of swabs (cloacal and/or oropharyngeal). The majority of birds (37/49) were admitted alive, prior to euthanasia or natural death, either to the Wildlife Care Center of the National Veterinary College of Alfort (EnvA; n = 36) or to a local diagnostic laboratory (Inovalys; n = 1). Upon admission, for all live birds, clinical assessments were performed, including evaluation of general condition, neurological status, and respiratory signs. The rest of the animals (12/49) were found dead in the field, collected, necropsied, and tested through the SAGIR network as part of the official HPAI national surveillance program34. Live gulls were humanely euthanized by intravenous injection of pentobarbital in the form of sodium salt (Dolethal ND), administered via the occipital venous sinus at a dose of 1.5 mL per bird. Selection of birds and pathology A complete necropsy was performed on all subjects. If the post-mortem assessment could not be conducted within 24 h of euthanasia or carcass retrieval, specimens were stored at -20 °C. From the original cohort of 49 black-headed gulls, 13 birds were selected for pathological characterization, including 8 euthanized individuals and 5 found dead. Selection was based on the degree of post-mortem preservation, including only specimens with minimal autolysis and negligible freezing artefacts, suitable for histopathological and immunohistochemical assessment. At necropsy, lesions were recorded and samples were collected in duplicate for molecular and histopathological analyses, including tracheal and cloacal swabs, immature primary flight feathers (if present), tissue sections of feathered skin (calvaria and tail region), nasal cavity, eyelid, eye, brain, trachea, lung, heart, skeletal muscle, liver, spleen, kidney, esophagus, proventriculus, gizzard, duodenum, ileum, caeca, gonads, and nasal, Harderian, adrenal, uropygial, and thyroid glands. For all cases, tissue sampling was performed either within 24 h post-mortem or after complete thawing of previously frozen carcasses. For histopathological assessment, tissues were fixed in 10% neutral buffered formalin immediately after sampling. Samples were shipped to the National Veterinary School of Toulouse (ENVT) at -20 °C (fresh frozen samples) or room temperature (formalin-fixed samples). Formalin-fixed tissue sections were routinely processed and embedded in paraffin blocks. Sections 3 μm thick were cut, mounted on glass slides, and stained with hematoxylin and eosin (H&E) stain. Slides were examined by light microscopy and lesions were recorded. Viral antigen tissue distribution To assess viral antigen tissue distribution, serial 3 µm-thick sections obtained from formalin-fixed and paraffin-embedded blocks of all collected tissues were mounted on charged slides for immunohistochemistry (IHC). The assay relied on a monoclonal mouse antibody directed against influenza A virus (IAV) nucleoprotein (NP) (Clone HB65, BE0159 Biozol, Eching, Germany), as previously described35. Immunohistochemical labeling was visualized following manufacturer’s recommendations, using 3,3'-diaminobenzidine (DAB) as the chromogen substrate (ENVISIO FLEX HRP Dako Omnis). Ocular sections were additionally stained using the Magenta chromogen system (ENVISO FLEX HRP Magenta system, Dako Omnis) to confirm positive viral antigen detection, and differentiate it from endogenous pigments. Sections of avian influenza virus (AIV) positive tissues were used as positive controls. Negative controls included sections incubated without the primary antibody or with a monoclonal antibody belonging to the same isotype (IgG2). Viral antigen tissue distribution was semi-quantitatively scored on a scale from 0 to 336. Molecular analysis Nucleic acids were extracted from all samples using the ID Gene Mag Fast Extraction Kit and IDEAL 32 extraction robot (Innovative Diagnostics, Grabels, France). The extracted RNA was screened by RT-qPCR targeting H5 and H7 hemagglutinin genes using the ID Gene Influenza H5/H7 Triplex PCR kit (Innovative Diagnostics, Grabels, France). Amplification and detection were performed on a LightCycler® 96 system (Roche Diagnostics, Basel, Switzerland). Sequencing analysis and phylogenetic analysis Whole genome sequencing was performed on RNA extracted from brain tissue of one PCR-positive black-headed gull using a metagenomic sequencing approach (Oxford Nanopore Technologies, UK)37. Briefly, reverse transcription and whole genome amplification were performed using influenza A–specific and universal primers38. Sequencing libraries were prepared using a Native Barcoding Kit and run on a MinION flow cell on a MinION Mk1C device37. Real-time basecalling was performed using Dorado integrated into MinKNOW. Phylogenetic analysis was performed on haemagglutinin (HA) gene sequences. Consensus sequences were aligned and compared with closely related reference sequences retrieved from BLAST searches and GISAID. Maximum-likelihood phylogenetic trees were inferred using IQ-TREE v1.6.12, applying the Hasegawa–Kishino–Yoneyama substitution model with gamma-distributed rate heterogeneity and 100,000 ultrafast bootstrap replicates. Trees were visualized using FigTree v1.4.2 (http://tree.bio.ed.ac.uk/software/figtree/). Results Clinical findings The 37 live gulls included 20 adults, 15 juveniles (first post-juvenile molt) and 2 subadults, comprising 17 males, 17 females and 3 individuals of undetermined sex. Affected birds were found immobile, weak and/or unable to fly, either in water or on land, across Île-de-France (Essonne, Seine-et-Marne, Val-de-Marne, Paris) and from Romorantin-Lanthenay (Loir-et-Cher) between 5 January and 6 February 2023. Clinical data were available for the 36 birds admitted to the Wildlife Care Center of EnvA. Overall, the most consistent and severe clinical signs observed included emaciation (25/36) and neurological signs (18/36). Neurological birds exhibited one or two of the following signs: altered alertness (10/18), head tremors (10/18), abnormal head posture (torticollis or opisthotonos) (9/18), pedaling of the hind limbs (3/18), which could be either permanent or intermittent, and jerky head movements (1/18) (Fig. 1a–f). Additional signs included persistent sternal recumbency (14/36), digestive (greenish diarrhea) (15/36), ocular (conjunctivitis with epiphora) (7/36) and respiratory disorders (wheezing or abnormal respiratory sounds) (7/36).Figure 1Clinical signs observed in black-headed gulls admitted to the Wildlife Care Center of the National Veterinary College of Alfort (EnvA). (a, b) Abnormal head posture with torticollis/opisthotonos. (c, d) Pedaling of the hind limbs. (e) Persistent sternal recumbency. (f) Soiled feathers with greenish droppings. Clinical signs observed in black-headed gulls admitted to the Wildlife Care Center of the National Veterinary College of Alfort (EnvA). (a, b) Abnormal head posture with torticollis/opisthotonos. (c, d) Pedaling of the hind limbs. (e) Persistent sternal recumbency. (f) Soiled feathers with greenish droppings. Necropsy findings The subset of gulls selected for pathological characterization included 10 adults, 2 juveniles (first post-juvenile molt) and 1 subadult. Necropsy reports were available for 8/13 cases, all involving euthanized birds. The majority of gulls were mildly to markedly emaciated (6/8) and only 2 individuals appeared in good nutritional status. Three birds exhibited moderately soiled plumage with greenish droppings. Significant visceral changes included mottled pancreas with pale areas and/or hemorrhagic foci (4/8), and multiple hemorrhages scattered in the mediastinum (2/8), meninges (1/8), periocular tissues (1/8) and pectoral muscles (1/8). Histopathology The most consistent lesions identified included mild to severe necrotizing and/or lymphoplasmacytic encephalitis (13/13), endophthalmitis (10/13), and pancreatitis (9/13). Mild necrotizing uropygial gland adenitis (5/7), necrosis of the feather epithelium and folliculitis involving immature contour feathers of calvaria skin sections. (1/13) were also present (Fig. 2).Figure 2Histopathological findings in black-headed gulls naturally infected with H5N1 HPAIV. (a) Cerebro-cortex: lymphocytes expanding Virchow-Robin spaces and infiltrating the neuroparenchyma (arrowhead) with concurrent gliosis and spongiosis, hematoxylin and eosin staining (H&E). (b) Choroid plexuses: multifocal segmental epithelial sloughing and necrosis (asterisk) and lymphocytic infiltration of underlying interstitium (arrowhead), H&E. (c) Ventricular ependyma: sloughing and necrosis of ependymocytes (arrowhead), H&E. (d) Eye, at low magnification: prominent non-suppurative infiltration of the irido-corneal angle (arrowhead), involving the trabecular meshwork, iris (asterisk), and ciliary bodies (arrow) H&E. (e) Eye: ciliary bodies exhibit non suppurative lymphocytic infiltration (arrowhead) and segmental sloughing and necrosis with pigmentary dispersion (asterisk), H&E. (f) Eye: segmental retinal necrosis and leukocytic infiltration (arrowhead), H&E. (g) Pancreas: extensive necrosis of exocrine acinar cells (asterisk), H&E. (h) Heart: discrete foci of cardiomyocyte necrosis with sarcoplasmic fragmentation and hypereosinophilia (arrowhead), H&E. (i) Lung: mild focal cellular densification of capillary beds with leukocytic infiltration, presumably lymphocytes (arrowhead), H&E. (j) Immature feather: epithelial sloughing, necrosis and loss of differentiating feather epithelium (arrowhead), H&E. (k) Uropygial gland: multifocal segmental necrosis of parenchymal cells (asterisks) engaged in holocrine secretion (arrowhead), H&E. (l) Eyelid: quill mites (Syringophilidae) expanding the quill and compressing the feather follicle epithelium of multiple periocular feathers (arrowhead), H&E. (m) Lung: pulmonary parenchyma exhibiting marked necro-inflammatory changes associated with numerous intralesional nematode eggs (arrowhead) and bacterial colonies (asterisk). Lesions include congestion, hemorrhages, fibrino-heterophilic exudation and necrotizing vasculitis (arrow), H&E. (n) Esophagus: focal parietal granuloma. The insert shows a nematode egg found on the mucosal surface, H&E. (o) Duodenum: nematode section surrounded by sloughed epithelial cells within the duodenal lumen, H&E. Scale bars: 50 µm (c, f, g, h and i); 100 µm (a, b, k, m, n and o); 200 µm (j and l); 500 µm (d). Histopathological findings in black-headed gulls naturally infected with H5N1 HPAIV. (a) Cerebro-cortex: lymphocytes expanding Virchow-Robin spaces and infiltrating the neuroparenchyma (arrowhead) with concurrent gliosis and spongiosis, hematoxylin and eosin staining (H&E). (b) Choroid plexuses: multifocal segmental epithelial sloughing and necrosis (asterisk) and lymphocytic infiltration of underlying interstitium (arrowhead), H&E. (c) Ventricular ependyma: sloughing and necrosis of ependymocytes (arrowhead), H&E. (d) Eye, at low magnification: prominent non-suppurative infiltration of the irido-corneal angle (arrowhead), involving the trabecular meshwork, iris (asterisk), and ciliary bodies (arrow) H&E. (e) Eye: ciliary bodies exhibit non suppurative lymphocytic infiltration (arrowhead) and segmental sloughing and necrosis with pigmentary dispersion (asterisk), H&E. (f) Eye: segmental retinal necrosis and leukocytic infiltration (arrowhead), H&E. (g) Pancreas: extensive necrosis of exocrine acinar cells (asterisk), H&E. (h) Heart: discrete foci of cardiomyocyte necrosis with sarcoplasmic fragmentation and hypereosinophilia (arrowhead), H&E. (i) Lung: mild focal cellular densification of capillary beds with leukocytic infiltration, presumably lymphocytes (arrowhead), H&E. (j) Immature feather: epithelial sloughing, necrosis and loss of differentiating feather epithelium (arrowhead), H&E. (k) Uropygial gland: multifocal segmental necrosis of parenchymal cells (asterisks) engaged in holocrine secretion (arrowhead), H&E. (l) Eyelid: quill mites (Syringophilidae) expanding the quill and compressing the feather follicle epithelium of multiple periocular feathers (arrowhead), H&E. (m) Lung: pulmonary parenchyma exhibiting marked necro-inflammatory changes associated with numerous intralesional nematode eggs (arrowhead) and bacterial colonies (asterisk). Lesions include congestion, hemorrhages, fibrino-heterophilic exudation and necrotizing vasculitis (arrow), H&E. (n) Esophagus: focal parietal granuloma. The insert shows a nematode egg found on the mucosal surface, H&E. (o) Duodenum: nematode section surrounded by sloughed epithelial cells within the duodenal lumen, H&E. Scale bars: 50 µm (c, f, g, h and i); 100 µm (a, b, k, m, n and o); 200 µm (j and l); 500 µm (d). Within the central nervous system, there was multifocal lymphocytic perivascular cuffing accompanied by gliosis, neuronal necrosis, and satellitosis, mostly in the parenchyma of cerebro-cortical hemispheres (Fig. 2a). Additionally, the ventricular system revealed sloughing and necrosis of ependymocytes, and choroid plexuses were occasionally expanded by lymphoplasmacytic cells (Fig. 2b,c). Ocular sections showed lymphoplasmacytic infiltration and necrotic degenerative changes involving the irido-corneal angle, trabecular meshwork of the pectinate ligament and ciliary bodies (Fig. 2d,e). Multifocal segmental degenerative/necrotic changes were occasionally noticed in retinal ganglion cells (Fig. 2f). Pancreatic acinar cells exhibited areas of acute necrosis with minimal inflammatory changes (Fig. 2g). Discrete necrosis/degeneration of cardiomyocytes was present in one bird (Fig. 2h), while a mild increase of cellularity was occasionally noticed in the pulmonary capillary bed (3/13) (Fig. 2i). In the integumentary system, the differentiating feather epithelium of immature contour feathers appeared necrotic (Fig. 2j). Uropygial gland sections presented with multifocal, segmental epithelial necrosis of the acinar epithelium, mostly involving the intermediate and outer cell layers, with basal cells largely preserved (Fig. 2k). Additional findings included moderate acariasis associated with quill mites (Syringophilida) involving the periocular (eyelid) feathers (2/13), severe necrotizing and exudative broncho-pneumonia associated with large numbers of intralesional bacteria and nematode eggs (1/13), a focal esophageal granuloma (1/13) and mild intestinal nematodiasis (1/13) (Fig. 2l–o). Immunohistochemistry Results of viral antigen tissue distribution, providing an overview of the visceral involvement, are summarized in Fig. 3. A detailed overview of immunolabelling of the central nervous system (CNS) and the eye, including the involvement of different anatomical compartments, is shown in Figs. 4 and 5, respectively.Figure 3Viral antigen distribution in black-headed gulls naturally infected with H5N1 HPAIV. (A) Viral antigen scores for the different organs analyzed from selected birds (n = 13), including the mean ± SD and frequency of positive detection. NA = not available. (B) Schematic representation of a gull with immunopositive organs highlighted in red. Representative images of viral antigen detection by immunohistochemistry in (a) brain, (b) eye, (c) pancreas, (d) uropygial gland, (e) lung, (f) feathered skin, (g) visceral ganglia and (h) adrenal gland. Scale bars: 25 µm (h); 50 µm (a, c, d, e and g); 100 µm (b and f). Schematic representation of visceral tropism in infected gulls (B) created with BioRender.com (Gaide, N., 2026; https://BioRender.com/2qtlyyz). *Others = eyelid, nasal cavity, trachea, heart, spleen, esophagus, proventriculus, gizzard, duodenum, ileum, caeca, gonad, nasal, Harderian gland and thyroid gland.Figure 4Viral antigen distribution in the central nervous system (CNS) of black-headed gulls naturally infected with H5N1 HPAIV. (A) Viral antigen scores for the different CNS compartments analyzed from selected birds (n = 13), including the mean ± SD and frequency of positive detection. NA = not available. (B) Schematic representation of a brain illustrating all immunopositive anatomical regions. Representative images of viral antigen detection by immunohistochemistry in (a) cerebrum, (b) midbrain, (c) cerebellum, (d) brainstem, (e) ventricular system, (f) choroid plexuses and (g) leptomeninges. Scale bars: 50 µm (a, insert, b, d, e and f); 100 µm (c and g).Figure 5Viral antigen distribution in the eye of black-headed gulls naturally infected with H5N1 HPAIV. (A) Viral antigen scores for the different ocular compartments analyzed from selected birds (n = 13), showing mean ± SD and frequency of positive detection. NA = not available. (B) Representative images of viral antigen detection by immunohistochemistry in (a) eyelid conjunctiva, (b) cornea, (c) trabecular meshwork, (d) iris, (e) ciliary bodies, (f) pecten oculi and (g) retina. No immunolabelling was observed in (h) optic nerve nor its meningeal lining. Scale bars: 25 µm (d); 50 µm (a and f); 100 µm (b, c, g and h). Viral antigen distribution in black-headed gulls naturally infected with H5N1 HPAIV. (A) Viral antigen scores for the different organs analyzed from selected birds (n = 13), including the mean ± SD and frequency of positive detection. NA = not available. (B) Schematic representation of a gull with immunopositive organs highlighted in red. Representative images of viral antigen detection by immunohistochemistry in (a) brain, (b) eye, (c) pancreas, (d) uropygial gland, (e) lung, (f) feathered skin, (g) visceral ganglia and (h) adrenal gland. Scale bars: 25 µm (h); 50 µm (a, c, d, e and g); 100 µm (b and f). Schematic representation of visceral tropism in infected gulls (B) created with BioRender.com (Gaide, N., 2026; https://BioRender.com/2qtlyyz). *Others = eyelid, nasal cavity, trachea, heart, spleen, esophagus, proventriculus, gizzard, duodenum, ileum, caeca, gonad, nasal, Harderian gland and thyroid gland. Viral antigen distribution in the central nervous system (CNS) of black-headed gulls naturally infected with H5N1 HPAIV. (A) Viral antigen scores for the different CNS compartments analyzed from selected birds (n = 13), including the mean ± SD and frequency of positive detection. NA = not available. (B) Schematic representation of a brain illustrating all immunopositive anatomical regions. Representative images of viral antigen detection by immunohistochemistry in (a) cerebrum, (b) midbrain, (c) cerebellum, (d) brainstem, (e) ventricular system, (f) choroid plexuses and (g) leptomeninges. Scale bars: 50 µm (a, insert, b, d, e and f); 100 µm (c and g). Viral antigen distribution in the eye of black-headed gulls naturally infected with H5N1 HPAIV. (A) Viral antigen scores for the different ocular compartments analyzed from selected birds (n = 13), showing mean ± SD and frequency of positive detection. NA = not available. (B) Representative images of viral antigen detection by immunohistochemistry in (a) eyelid conjunctiva, (b) cornea, (c) trabecular meshwork, (d) iris, (e) ciliary bodies, (f) pecten oculi and (g) retina. No immunolabelling was observed in (h) optic nerve nor its meningeal lining. Scale bars: 25 µm (d); 50 µm (a and f); 100 µm (b, c, g and h). Viral antigen was most commonly detected in brain (13/13), eye (10/13) and pancreas (8/13), followed by uropygial gland (4/7), lung (3/13), feathered skin (2/13), visceral ganglia (2/13) and adrenal gland (1/9). The highest immunohistochemical scores were recorded in brain (2.2) and pancreas (2.1), followed by eye (1.8), and uropygial gland (1.5). Lung, feathered skin, visceral ganglia and adrenal gland had the lowest score (1) (Fig. 3A). No viral antigen was detected in any of the gastrointestinal tract segments examined (esophagus, proventriculus, gizzard, duodenum, ileum, and caeca). In terms of subcellular detection, viral antigen was noticed within the nucleus and the cytoplasm of parenchymal and inflammatory cells (Fig. 3B). There was no endothelial labelling in any of the tissues examined. Within the CNS, viral antigen was consistently and most abundantly detected in the cerebrum (13/13), followed by ventricular system (10/12) and midbrain (10/11). The cerebellum (5/9), brainstem (4/7), choroid plexuses (2/8), and leptomeninges (4/13) were less frequently involved and generally exhibited lower immunohistochemical scores (Fig. 4A). Antigen distribution within the CNS was mostly multifocal to coalescent and varied among individuals, with a predominance of neuronal and glial labeling (Fig. 4B). In the eye, viral antigen was most frequently detected in the retina (5/13) and trabecular meshwork (6/13), followed by cornea (4/13), iris (3/13), and ciliary body (3/13) (Fig. 5A). Sporadic labeling was observed in the conjunctiva and pecten oculi, whereas the choroid and optic nerve were consistently negative. Viral antigen distribution tended to be focal to multifocal, preferentially involving the anterior segments (Fig. 5B). Magenta staining confirmed standard immunohistochemical findings with DAB (Supplementary Fig. 1). Molecular analysis Results are detailed in Table 1. All birds were confirmed IAV H5 positive by RT-qPCR and viral RNA was detected in the vast majority of the organs examined. Specifically, the brain and pancreas exhibited the lowest Ct values (average CT values of 15 and 18) followed by lung, eye, Harderian and nasal glands, gonads and small intestine (average between 22 and 23). The highest Ct values were identified in cloacal swabs and feathers (average 31). The rest of the tissues showed average Ct values ranging between 23 and 29.Table 1Detection of H5 IAV RNA by RT-qPCR in black-headed gulls naturally infected with H5N1 HPAIV.Cycle threshold (Ct) values for tracheal and cloacal swabs and the different organs analyzed from selected birds (S1 to S13, n = 13), including the mean ± standard deviation (SD) and frequency (Freq) of positive detection. TR = tracheal swab. CL = cloacal swab. NA = tissue not available. Neg = negative. Table 1Detection of H5 IAV RNA by RT-qPCR in black-headed gulls naturally infected with H5N1 HPAIV.Cycle threshold (Ct) values for tracheal and cloacal swabs and the different organs analyzed from selected birds (S1 to S13, n = 13), including the mean ± standard deviation (SD) and frequency (Freq) of positive detection. TR = tracheal swab. CL = cloacal swab. NA = tissue not available. Neg = negative. Detection of H5 IAV RNA by RT-qPCR in black-headed gulls naturally infected with H5N1 HPAIV. Cycle threshold (Ct) values for tracheal and cloacal swabs and the different organs analyzed from selected birds (S1 to S13, n = 13), including the mean ± standard deviation (SD) and frequency (Freq) of positive detection. TR = tracheal swab. CL = cloacal swab. NA = tissue not available. Neg = negative. Sequencing and phylogenetic analysis Whole genome sequencing of brain tissue yielded a near-complete influenza A virus genome. Phylogenetic analysis of the haemagglutinin gene placed the virus within the H5N1 clade circulating in Europe in 2023, showing closest similarity to strains detected in the Netherlands (bootstrap support: 90%; Supplementary Figure S2). The sequences were deposited in GenBank (PX933532–PX933537). Discussion This study provides an in-depth pathological and molecular characterization of selected cases of HPAI H5N1 in wild black-headed gulls in France, with a particular focus on central nervous system (CNS) and ocular involvement. As previously reported in this species, the CNS was the most consistently and severely affected organ at histopathology, which likely accounted for the clinical signs observed. Similar patterns have been described in earlier studies of HPAI H5N1 in Charadriiformes, including black-headed gulls14,18,21,22,39. During a mortality event in the Netherlands, marked neurotropism together with the involvement of pancreas, lung, and uropygial gland was observed in black-headed gulls and common terns (Sterna hirundo)22. Similarly, a recent UK study characterized HPAI H5N1 neurotropism across several Charadriiformes species. However, inflammatory changes were not consistently observed, and detailed anatomical assessment of the CNS was limited to four black-headed gulls18. Our work provides a detailed anatomical mapping of IAV antigen distribution within the CNS of black-headed gulls, based on a relatively large case series compared to previous reports. Viral antigen was consistently detected in the cerebral hemispheres, with frequent involvement of the midbrain, ventricular system, and brainstem. Neurons and neuropil were invariably labelled, with immunolabelling commonly observed in the choroid plexus and leptomeninges. Consistent with previous studies, viral antigen was not identified in the endothelial cells of the CNS nor the rest of organs examined18,22. This species-specific characterization of CNS infection clarifies the neuroanatomical distribution of viral antigen and allows the identification of significant sites of viral replication in black-headed gulls. Recovering the entire CNS may not always be feasible in routine post-mortem evaluation, particularly when working with wildlife. However, from a research perspective, systematic sampling of the brain is essential to properly assess viral tropism and associated lesions, interpret clinical signs observed in live birds and correlate molecular and histopathological data. From this point of view, a portion of the rostral cerebrocortex (coronal section) should be collected for molecular analysis and stored at − 80 °C, while the remaining tissue should be fixed in formalin for histopathology. In the present study, we specifically decided to combine the analysis of CNS and ocular tissues, as these organs are anatomically and physiologically linked40. Infection of one organ may have implications for the other, providing key insights for understanding the pathophysiology of HPAI H5N1 and its short and long-term impact in affected birds. The high prevalence of ocular lesions in our set of birds, including endophthalmitis/anterior uveitis and retinal degeneration, combined with positive immunohistochemical and molecular detection, indicates that the eye is a target organ of HPAIV H5N1 virus in black-headed gulls. Pathogens may gain access to the eye via direct contact with the conjunctival mucosa or corneal epithelium, through contaminated dust, water or fomites, leading to conjunctivitis and keratitis. This route was used in experimental studies of HPAIV H5N1 infection in common gulls (Larus canus)41. Microorganisms colonizing the nasal cavity and oropharynx may also reach the eye through the nasolacrimal ducts and the choana, connecting the upper respiratory and digestive tract to the ocular surface42. Intraocular structures may be invaded hematogenously in the event of a systemic infection, potentially leading to uveitis, endophthalmitis, and retinitis22. Concurrent corneal opacity, resulting from acute edema and keratitis, is reported in several waterfowl species, and laughing gulls (Leucophaeus atricilla) inoculated with intranasal HPAIV H5N143. The eye is a highly vascularized organ, with the uvea, including the iris, ciliary body and choroid, playing a central role in ocular physiology. The ciliary body is involved in the production of the aqueous humor, which represents a plasma ultrafiltrate actively modified by the non-pigmented epithelial cells of the ciliary processes. It maintains intraocular pressure and supplies oxygen and nutrients to the avascular tissues of the anterior segments44,45. Although direct evidence of HPAIV vascular invasion of intraocular structures in birds is lacking, these anatomical and physiological features make hematogenous infection plausible. Pathogens may also reach the eye via the CNS, traveling along the optic nerve, from infected cerebral tissue to the retina and other intraocular structures46. In mammals, ocular exposure can also lead to respiratory and systemic infection, although the involvement of intraocular structures remains unclear. In swine, ocular inoculation with H1N1 (A/swine/Guangdong/1/2011) resulted in viral replication in the conjunctiva and spread to the respiratory tract via the nasolacrimal duct, but no ocular lesions were observed47. In mice, ocular exposure to IAV H5 and H7 subtypes resulted in viral replication in the corneal epithelium and subsequent systemic infection, including the brain. However, detailed examination of intraocular structures was not performed48. In ferrets, ocular inoculation of HPAIV H5N1 led to systemic infection and viral RNA and infectious virus were recovered from conjunctival washes during the acute phase of disease. Viral antigen was also identified in the epithelial cells of the lacrimal glands and ciliary processes49. Overall, the visceral lesions observed in our study were indicative of a systemic infection, suggesting hematogenous viral dissemination. The frequent and severe involvement of the irido-corneal angle, further supports vascular spread of AIV to the ocular compartment. Conjunctival immunostaining was very limited, while spread through the optic nerve was considered unlikely, since no histopathological lesions or viral antigen were detected at this level. Nevertheless, in birds, the optic nerve is surrounded by meningeal sheaths continuous with the intracranial meninges, which could provide an additional pathway for viral entry into the eye50. HPAI has caused significant mortality in a variety of marine bird species, including some that are endangered5–11. However, carcasses recovered in the field, may represent only the tip of the iceberg. In wintering marine birds, reduced daylight, adverse weather conditions, and variable food availability and quality, can increase energy demands and results into significant inter-individual variation in body condition51. Juveniles may be particularly affected due reduced foraging and competitive abilities associated with inexperience. This seasonal context may partly explain the heterogeneity in the nutritional status observed in our study, together with the rapid deterioration associated with acute HPAIV infection. Beyond acute mortality, some individuals may survive but develop more chronic lesions, that could reduce their chances of long-term survival. A recent study in snow geese (Anser caerulescens) naturally-infected with H5N1 HPAIV suggests that persistent neurological lesions may adversely affect migratory behaviour52. In addition to neurological involvement, HPAIV infection has been associated with ocular abnormalities, such us “black eyes”, in wild marine birds, although these findings remain less consistently identified and characterized across species16,17. The pathophysiology of these changes was not investigated, however, at least three potential mechanisms, acting alone or in combination, may be involved: impaired aqueous humor outflow, altered iris muscular function and pigmentary incontinence. Chronic anterior uveitis or iridocyclitis can induce the formation of synechiae between the anterior surface of the iris and the trabecular meshwork, reducing aqueous humor outflow and leading to increased intraocular pressure, secondary glaucoma and structural iris changes53. This process resembles uveitic glaucoma, a form of secondary glaucoma described in humans following numerous viral infections54. Based on our findings, compromised aqueous humor drainage represents a plausible mechanism. Impaired neuromuscular control of the iris may also contribute to the development of “black eyes”. Unlike mammals, the avian iris contains predominantly striated muscle fibers, which regulate pupil size under somatic motor control, primarily via the oculomotor nerve53. Damage to the iris sphincter muscle or its innervation can result in persistent mydriasis, causing the pupil to occupy a large proportion of the visible iris, and giving the appearance of darkened or “black” iris, despite the absence of pigmentary changes. In birds, the pupillary light reflex is mediated by retinal ganglion cells whose axons travel through the optic nerve to the midbrain pretectal nuclei. These nuclei project to neurons in the Edinger–Westphal complex. From there, efferent fibers travel through the oculomotor nerve to the ciliary ganglion, and postganglionic neurons ultimately innervate the iris sphincter, completing the reflex pathway55,56. Damage to any portion of this neuronal pathway could impair pupillary constriction. Considering the neurotropism and significant involvement of midbrain observed in our cases, a similar mechanism cannot be excluded. Finally, necrotic-inflammatory changes affecting pigmented iris cells may lead to pigmentary incontinence. This process is characterized by the release and extracellular dispersion of melanin granules, producing patchy or mottled irises57. Overall, these mechanisms remain hypothetical and are based on indirect pathological observations rather than experimental or functional evidence. In contrast to northern gannets16,17, gulls in our study did not show abnormally-colored irises or anisocoria. However, since most birds were euthanized shortly after clinical assessment, we could not exclude the possibility that similar changes might have developed had the birds survived. A recent study in northern gannets reported that individuals exhibiting “black eyes”, indicative of prior HPAIV infection, showed no detectable differences in foraging behaviour or reproductive success compared with birds with normal eye colour17. Even during the acute phase of disease, ocular lesions could impair vision, particularly in bright light, and reduce depth perception and predator detection. Neurological signs, such as ataxia, tremors, and weakness likely exacerbate these deficits, further affecting foraging and social interactions. Chronic ocular changes, including persistent mydriasis, along with CNS lesions, may also have long-term effects on vision, coordination, balance and behavior, negatively affecting the survival of birds recovering from acute infection. Consistent with the increased risk of predation, several peregrine falcons (Falco peregrinus) were found dead in areas affected by black-headed gulls’ mortalities in France, suggesting potential viral transmission through active predation of infected and/or sick gulls or scavenging of carcasses4,39,58–60. Additional studies are needed across different species and ecological contexts to evaluate both the short-term and long-term impact of HPAI-induced ocular and neurological changes on overall fitness and their implications for HPAI ecology and transmission. Iris abnormalities could represent a non-invasive indicator of prior HPAIV infection16,17. However, proper interpretation requires careful consideration of species-specific, age-related, and hormonal variation in iris pigmentation, as well as pupil size and symmetry. Our results support the potential role of the integumentary system as an additional shedding route for AIV in marine birds. The vast majority of birds examined were adults, as indicated by the paucity of immature feathers, particularly in the calvarium and tail region. However, when growing feathers were present, epithelial necrosis and viral antigen detection were identified, supporting feather epitheliotropism, as previously described in domestic ducks35,61. Viral antigen was also identified in degenerated/necrotic acinar epithelial cells in the uropygial gland. The involvement of the intermediate and outer layers of acinar cells, undergoing terminal disintegration as part of the holocrine secretion process62, is similar to the pattern of viral antigen detection in the outer and intermediate layers of the feather epithelium, comprised of degenerated and keratinized epithelial cells in ducks61. Integumentary infection could contribute to viral dissemination within colonies, including birds of different ages, ranging from hatchlings covered in soft down, chicks with emerging feathers to fledgling juveniles. In addition, during molting, birds are temporarily flightless and tend to gather at specific sites, mixing with individuals of different ages, which creates favorable conditions for HPAIV transmission and persistence prior to fall migration63. These findings, combined with our results, highlight the importance of multidisciplinary studies, integrating pathology, ecology and movement analysis, to identify high-risk periods and locations for viral spread and ultimately to improve our understanding of the epidemiology of HPAIVs in wild birds. It should be noted that the observational nature of this study, based on a limited number of spontaneous cases of HPAIV H5N1 infection in wild black-headed gulls, represents an important limitation, and the findings may not be fully generalizable to all HPAI H5N1 strains, epidemiological contexts, or infection conditions. Conclusion In the present study, HPAIV H5N1 infection in black-headed gulls appears predominantly neurotropic, oculotropic, and selectively epitheliotropic, with no evidence of endotheliotropism. This tissue tropism closely resembles that described in domestic ducks and may have important intra- and interspecific implications for viral diffusion, persistence, and transmission. However, there are still many questions on how this tropism affects patterns of infection, host adaptation, and the maintenance of HPAI viruses in wild bird reservoirs. Our findings underscore the value of including the brain, eyes, and integumentary system in HPAI surveillance and diagnostic protocols in marine bird species, to assess the role of these compartments in viral ecology and transmission. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3.