Fig 2: Probability for rupture of MUC1-Tn and MUC1-STn MGL interactions with MGL prior to and after the addition of free Tn-Ser or STn-Ser, respectively, to the buffer solution.A: Fraction of total number of AFM force—distance curves displaying signatures of intermolecular rupture events (Fint), prior to (black) and after (grey) the addition of Tn-Ser or STn-Ser to a concentration equal to 0.5 mg/ml. B: Cumulative fraction of work performed during the forced unbinding of MUC1-Tn (continuous lines) and MUC1-Tn (dotted lines) from MGL. The analysis is performed based on the fraction of the AFM force-distance curves where interaction events were observed. The total deadhesion work determined both prior to (black lines) and after (grey lines) adding Tn-Ser or STn-Ser are displayed. The symbols represent the mean de-adhesion work performed on the MUC1—MGL molecular pairs (Eq (2)) and are located at corresponding percentiles in the cumulative distributions.

Fig 3: MUC1-Tn and STn show a similar and density dependent deadhesion work when allowed to interact with MGL, whereas MUC1-ST show a low and density independent deadhesion work.A: Fraction of total number of AFM force—distance curves displaying signatures of interaction (Pint). The Pint values determined based on curves obtained using AFM tips functionalised with a low or a high density of glycans are presented separately. B: Distribution of work performed on the MUC1-Tn-MGL (blue), MUC1-STn–MGL (green) and MUC1-ST–MGL (red) molecular interactions in forced unbinding experiments. performed using an AFM tip retraction speed equal to 2 µm/s. The histogram distributions are based on the following number of observations: MUC1-Tn: 339, MUC1-STn: 283 and MUC1-ST: 62. C: Cumulative fraction of the distributions presented in panel B of work performed on the MUC1–MGL molecular pairs during the forced unbinding. The colour codes are the same as in B. The symbols represent the mean de-adhesion work performed on the MUC1-MGL molecular pairs (Eq (2)) and are located at corresponding percentiles in the cumulative distributions.

Fig 4: Cells expressing MGL bind to MUC1-STn glycopeptides and glycoproteins in a calcium dependant manner.Immature moDC were generated as described in materials and methods and incubated with fluorescent beads pre-coated with A, recombinant MUC1 glycoproteins; B, MUC1 glycopeptides. Cell binding was assessed by flow cytometry, with the % of binding cells shown. One experiment representative of three independent experiments is shown. C, MoDC were matured with LPS and the binding of MUC1 glycopeptides determined as above. D, MGL expression levels on immature and mature monocyte derived DCs, black; isotype control: bold black; anti-MGL binding. E, K562 (i,iii) and K562 transfected with MGL (ii,iv) were incubated with fluorescent beads pre-coated with recombinant MUC1 glycopeptides (i,ii) or recombinant MUC1 glycoproteins (iii,iv). Cell binding was assessed by flow cytometry, with the % of binding cells shown. One experiment representative of two independent experiments is shown. Inserts show the expression of MGL for K562 cells and K562 cells transfected with MGL, black; isotype control: bold black; anti-MGL binding.

Fig 5: T47D breast carcinoma cells expressing either STn or Tn, can bind to K562-MGL cells in a calcium and time dependant manner.A, dot-plots illustrating the interaction between T47D-STn cells and K562 or K562-MGL cells in the presence or absence of calcium. T47D-STn and K562 cells were stained with dyes (eFluor 670 and CFSE respectively) before being incubated on ice together for 4h. Cell:cell interaction was measured by flow cytometry as double positives. Cells were gated using CFSE and FSC. B, graphic representation of figure A with the addition of T47D-Tn binding to K562-MGL. C, measuring this interaction at different time points reveals the peak interaction occurs between 4 and 24h for both T47D-Tn and T47D-STn cells.