We included 2 archived bovine mammary glands obtained from the same lactating dairy cow (4 years of age) of a Danish Holstein breed and included mammary glands from 8 cows of different breeds and ages obtained from a slaughterhouse in Denmark (Table). The tissues were formalin-fixed, paraffin-embedded, and cut into 4–5 µm sections.

Figure 1

Figure 1. Results of staining showing wide expression of the human receptor for influenza A virus (designated SNA) in the bovine mammary gland. A) An example of the SNA staining of a...

Figure 2

Figure 2. Results of staining showing wide expression of the duck receptor for influenza A virus (designated MAA-II) in the alveoli of the bovine mammary gland. A) An example of the MAA-II...

We detected the human receptor SA-α2,6 using biotinylated Sambucus nigra lectin (SNA) (B-1305-2; Vector Laboratories, https://vectorlabs.com). We detected the chicken receptor SA-α2,3-Gal-β1,4 using biotinylated Maackia amurensis lectin I (MAA-I) (B-1315-2; Vector Laboratories) and the duck receptor (SA-α2,3-Gal-β1,3) biotinylated MAA-II (B-1265-1; Vector Laboratories) as previously described (13). We semiquantified the staining on the surface of the epithelial cells as the percentage of positive surface area in 2 images from each slide (Figure 1, panels D, E; Figure 2, panels D, E) and reported the average (13). We applied 2 negative controls. To investigate potential background staining (blind), we added no lectins; to investigate the amount of nonspecific binding, we applied a neuraminidase pretreatment before each lectin, as previously reported (13). We included a confirmed IAV-negative pig lung as a positive control.

Figure 3

Figure 3. Pearson correlation coefficient showing a significant negative correlation between the percentage of SNA staining (bovine receptor) in the alveoli and the parity status of the cows (A) but not for...

In the mammary gland, the human receptor (detected by SNA) and the duck receptor (detected by MAA-II) were widely distributed in the alveoli, but less so in the ducts, whereas we detected no positive staining of the chicken receptor by MAA-I (Table; Figure 1; Figure 2; Appendix Tables), except for 1 cow expressing the chicken receptor in a lactiferous duct. Both human and duck receptors were primarily expressed in the active alveoli and less expressed in less active alveoli (Figure 1; Figure 2). We found a significant negative correlation between the percentage of SNA staining in the alveoli and the parity status of the cows but not with the MAA-II lectin staining (Figure 3).

We detected nonspecific staining (positive staining after neuraminidase pretreatment) in the endothelial cells with the SNA and MAA-I lectins, in the bovine erythrocytes with the MAA-I lectin, and in the ducts of the MAA-II lectin, but observed no nonspecific staining elsewhere (Appendix Figure). The positive controls of the pig lung corresponded to previous findings (13) (Appendix Figure).

Our investigation evaluated the expression of IAV receptors in situ in the mammary gland of cattle, which typically has been considered less susceptible to IAV infection (2). Of note was the finding that both the human receptors (SA-α2,6) and the duck receptors (SA-α2,3-Gal-β1,3) were highly expressed in the active alveoli in mammary glands but lower expressed in the less active alveoli, which could indicate that cows peaking in lactation are more susceptible for infection. The findings of the 2 receptors are in agreement with 2 novel studies investigating the IAV receptors in the bovine mammary gland of 2–3 cows (11; M.R. Carrasco et al., unpub. data, https://doi.org/10.1101/2024.05.24.595667). The transmission routes and the pathogenesis of H5N1 in cows remain unclear, but the virus remains infectious after 1 hour on the milking equipment (V.L. Sage et al., unpub. data, https://doi.org/10.1101/2024.05.22.24307745); the US Department of Agriculture has reported that only some udder quarters may be involved in infection, suggesting an ascending infection as a possible transmission route. Of interest, the human and duck receptors were less expressed in the ducts of the mammary gland, making an ascending mammary gland infection using those receptors more challenging. An investigation of the receptor binding preference of A/Texas/37/2024 showed that an I199T mutation of the hemagglutinin protein increased the receptor binding breadth to a higher number of conformations of the avian receptor, including hybrid N-glycans, compared with previous H5N1 viruses (M.R. Good et al., unpub. data, https://doi.org/10.1101/2024.06.22.600211); those hybrid N-glycans might not be detected by the lectins used in our study (14). M.R. Carrasco, et al. (unpub. data) showed limited binding of older AIV strains and a mouse-adapted human influenza strain (PR8) in the mammary gland of cows, whereas the preliminary report from FLI (6) showed that genotypes from Europe of the clade 2.3.4.4b H5 HPAI viruses could also replicate in the mammary glands of cows. Thus, more studies are needed to investigate the susceptibility of cows to different IAV strains and variants.

The SNA lectin detects both of the 2 most common sialic acids, N-acetylneuraminic acid (Neu5Ac) and N-glycolylneuraminic acid (Neu5Gc) (14). Neu5Ac is mainly the preferred SA for IAVs, except for equine IAVs, which have a higher preference for Neu5Gc (7). Cattle express both Neu5Ac and Neu5Gc in their tissues, but the amount of Neu5Ac in bovine milk is 30 times higher than for Neu5Gc (15), which indicates that the SNA staining detected in our study was caused by staining of Neu5Ac. However, further studies such as mass spectrometry, which gives detailed information about the IAV receptors (e.g., sialic acid type, length, branching), are needed to confirm the cause of SNA staining and also to accomplish a comprehensive quantification of the distribution of receptors in bovine tissues (8).

The expression of the duck receptor in the mammary gland of cows fits with the observed widespread infections among cattle in the United States with HPAI H5N1. The co-expression of both human and avian receptors in the mammary glands indicates susceptibility to other IAVs than those from avian origin, which is worrying from a zoonotic perspective. The presence of the IAV receptors, however, does not provide evidence that cattle are susceptible to all IAVs; other host factors (1) probably play a role for successful replication.