Glycan distribution and density in native skin's stratum corneum

Abstract Background The glycosylation of proteins on the surface of corneocytes is believed to play an important role in cellular adhesion in the stratum corneum (SC) of human skin. Mapping with accuracy the localization of glycans on the surface of corneocytes through traditional methods of immunohistochemistry and electron microscopy remains a challenging task as both approaches lack enough resolution or need to be performed in high vacuum conditions. Materials and methods We used an advanced mode of atomic force microscope (AFM), with simultaneous topography and recognition imaging to investigate the distribution of glycans on native (no chemical preparation) stripped samples of human SC. The AFM cantilever tips were functionalized with anti‐heparan sulfate antibody and the lectin wheat germ agglutinin (WGA) which binds specifically to N‐acetyl glucosamine and sialic acid. Results From the recognition imaging, we observed the presence of the sulfated glycosaminoglycan, heparan sulfate, and the glycans recognized by WGA on the surface of SC corneocytes in their native state. These glycans were found associated with bead‐like domains which represent corneodesmosomes in the SC layers. Glycan density was calculated to be ~1200 molecules/μm2 in lower layers of SC compared to an important decrease, (~106 molecules/μm2) closer to the surface due probably to corneodesmosome degradation. Conclusion Glycan spatial distribution and degradation is first observed on the surface of SC in native conditions and at high resolution. The method used can be extended to precisely localize the presence of other macromolecules on the surface of skin or other tissues where the maintenance of its native state is required.


| INTRODUCTION
The glycosylation of cell membrane proteins plays an important role in cell-cell interactions, cell adhesion, proliferation, differentiation, morphogenesis, remodeling of the extracellular matrix, hydration, antimicrobial activity, 1,2 and modulation of inflammatory responses. 3 In human skin, N-linked glycosylation is the most common type of post translational modification on Asn-X-Ser/Thr motives of epidermal proteins. Here N-glycans contain a common core pentasaccharide known as the trimannosyl core which can be fucosylated on the first N-acetyl glucosamine (core fuscosylation). N-linked glycans are branched oligosaccharides consisting of 3 types: High mannose/oligomannoseconsisting, as the name suggests, only of branched mannose chains; Hybrid-addition of N-acetyl glucosamine, galactose, or fucose to the trimannosyl core or oligomannose structures; Complex-the addition of terminal sialic acids to hybrid N-glycan structures. 4 A second major glycosylation found on the surface of keratinocytes is the covalent O-linkage of the sulfated glycosaminoglycan heparan sulfate to Ser/Thr residues of proteins. Heparan sulfated proteoglycans (HSPGs) and N-linked glycosylated proteins have been postulated to play a role in proliferation and differentiation of the epidermis. 5,6 In the stratum corneum of human skin (Figure 1), protein glycosylation has been postulated to play an important role in the desquamation process, the shedding of corneocytes from the surface of the skin, since glycan moieties have been demonstrated to protect the corneodesmosomes from proteolysis. 7 There has been a paucity of information regarding the presence of heparan sulfate or N-linked glycosylation in the stratum corneum. A histological study of skin sections using plant lectins demonstrated the presence of glyco-conjugates in the stratum corneum but at a significantly lower level as compared to that in epidermis. 8 A glycomic study confirmed the presence and changes of N-linked glycans in the stratum corneum in the dry skin condition, ichthyosis. 9 A more recent electron microscopic study using plant lectins to detect N-glycans and an antibody to heparan sulfate evidenced the distribution of these glycans on the surface of delipidated corneocytes. 10 In these studies, the localization of glycans on the surface of corneocytes through the traditional methods of immunohistochemistry and electron microscopy has required the removal of the lipid-rich layers of the stratum corneum and may not present a true distribution of these glycans on the SC surface. Thus, in order to obtain a more realistic picture of glycan localization on the surface of native corneocytes, we used an advanced AFM TREC 11,12 mode with AFM cantilever tips functionalized with anti-heparan sulfate antibody and WGA, respectively to detect the presence of heparan sulfate and complex N-glycan chains on the SC surface. This method was proved previously to be well adapted for the study of this skin surface layer. 13

| Force measurements
The presence of heparan sulfate on the surface on corneocytes was investigated on stripped abdominal SC samples using single molecular recognition force spectroscopy, 14 for which the cantilever tips were functionalized with an anti-heparan sulfate antibody ( Figure 2). In this method, the interaction force between the tip and the surface of the corneocyte is measured. An example of force distance curve is shown in Figure 3. The observed binding probabilities were around 15%. The specificity was checked by injecting a solution with free anti-heparan sulfate antibodies resulting in more than threefold drop of the binding probability. From the force distance curve measurements, mostly single rupture events were observed with the anti-heparan sulfate antibody. Occasionally, a multistep rupture was obtained. Such events were interpreted as serial unbinding of the individual antibody arms.
The distributions of the rupture forces showed a bimodal behavior ( Figure 4), which were fitted with a sum of 2 Gaussians. The Gaussian located at lower forces was attributed to rupture events of a single antibody arm. Accordingly, the Gaussian located at higher forces was interpreted as a simultaneous rupture of both antibody arms. The most probable rupture force of the single antibody rupture showed a linear rise with respect to the logarithm of the loading rate. These data were fitted using Evan's single energy barrier model 15,16 where F* denotes the most probable rupture force, r the loading rate, k off the thermal off rate, k B the Boltzmann constant, T the temperature, and x β the thermally averaged length of interaction along the with the uncorrelated Markov model using the above acquired parameters. This represents a configuration for multiple-bond attachments, where load is shared between all bonds (parallel attachment).
Rupture dynamics of multiple bonds depends on the failure mode. In the correlated mode, all bonds are closely coupled, and failure of 1 bond implies failure of remaining bonds. In the uncorrelated system, the attachments can fail independently and the load force is redistributed among surviving bonds. Since both arms of the antibody do not cause close coupling, the analysis for uncorrelated failure was used. Uncorrelated failure mode implies no particular mechanical coupling between individual bonds; it can therefore be described as a Markovian sequence. 17 The measured unbinding force, ie, most probable rupture force F* scales with the number of bonds, NB and the measured loading rate, r according to: where x β and k off denote the corresponding parameters derived from a single bond analysis using Equation 1. Figure 5 shows this analysis applied to the full force spectroscopy experiment of single antibody arm and full antibody dissociation. The single arm interaction was fitted using Equation 1, yielding x β and k off . These parameters were used together with Equation 2 to calculate the dynamic force spectrum for full antibody interaction. Both calculation models adequately agree with the obtained data. Our force spectroscopy data provide the proof of principle for the detection of molecular recognition, in which the antibody bound specifically to the heparan sulfate with high affinity. Thus, the functionalized AFM tip is an appropriate tool for localizing recognition sites. The finding of double-arm binding of the antibody indicates a high surface density of heparan sulfates.
Schematic illustration of the functionalization of the cantilever tip with antibody or WGA. The AFM cantilever tip is first treated with 3-aminopropyl-triethoxysilane (APTES) to obtain the amine groups (see Methods) which can react with the NHS group of the PEG linker in the presence of triethylamine. After application of citric acid, the acetal group of the linker converts to aldehyde that subsequently reacts with the amine group of the lysine residue of the anti-heparan sulfate antibody or WGA. Since both antibody and WGA are coupled to the tip through flexible linkers, they have the possibility to adjust to their ideal orientation for binding to the glycan. In the presence of sodium cyanoborohydride (NaCNBH 3 ), the proteins are irreversibly linked to the AFM tip. Their density is sufficiently low to allow for single molecule studies [Colour figure can be viewed at wileyonlinelibrary.com] F I G U R E 3 When approaching the cantilever tip, (blue curve), the distance between the tip and the cell surface decreases. At the contact point, the tip touches the cell membrane. After the contact (left side of the contact point), the tip gently presses on the cell membrane. When the force limit (which is about 50 pN here) is reached, the tip is retracted from the cell surface. If the antibody binds to the heparan sulfate, the cantilever tip is pulled downwards (red curve) until the 2 molecules are separated at a critical force. After the rupture, the cantilever recovers its resting state. If there is no binding between antibody and heparan sulfate during the contact, the retraction curve looks similar to the approaching curve [Colour figure can be viewed at wileyonlinelibrary.com]

| Localization of heparan sulfate and WGAbinding glycans on corneocytes
Force spectroscopy allowed specific glycans to being detected and identified on the corneocyte surface, yet information about the exact position of the glycans or relation of the glycans with the topography is missing.
Combining the topography information of imaging with the identification capability of force spectroscopy allows for specific localization of glycans with corresponding surface features. This is made possible by the simultaneous topography and recognition imaging technique. 11,13 At first, a stripped abdominal SC sample was investigated to localize the heparan sulfate on the surface. For this, antibodies against heparan sulfate were tethered to AFM tips as described in the Methods section. The tip used had a spring constant of 0.1 N/m, which was 3-10 times higher than that of tips used for force spectroscopy. The resulting resonance frequency is about 7.5 kHz in solution and the Qfactor is about 1. The oscillation amplitude was adjusted to be less than the extended PEG linker to provide a proper recognition image. 18 As a result, the recognition map represents an amplitude reduction due to a physical connection between the heparan sulfate antibody on the tip and heparan sulfate molecules on the skin sample surface when specific interaction occurs. These recognition spots are nonuniformly distributed as microdomains with dimensions from several nanometers up to 100 nanometers ( Figure 6C, red spots). The recognition events correspond with beads observed in the topography image.
Repeated scans of the same area showed that the recognition was stable. A threshold analysis was used to identify recognition events.
The threshold was set to be the mean value minus 1.414 × SD of this image. Taking  Moreover, using TREC we correlated the recognized glycan sites with the bead-like structure of corneodesmosomes in the SC.
It has been shown that both peripheral and non-peripheral corneodesmosomes persist at the corneocyte surface in the upper SC layers of xerotic skin. 19 Thus, AFM is a very useful tool to characterize the morphological changes and the distribution of other corneodesmosomal components (eg, desmoglein and desmocolin) and various glycan moieties (eg, hyaluronan) on the surface of corneocytes. Such information will aid designing new skincare-compounds prone to better interact with the SC surface.  was dialyzed in 250 mL PBS (PAA) 2 times to remove sodium azide.

| AFM experiments
All TREC measurements were taken using a PicoPlus 5500 AFM were performed at room temperature using antibodies coupled with AFM tips with 0.01-0.03 N⁄m nominal spring constants. The cantilever spring constants were determined using the thermal noise method. 29,30 Empirical force distribution of the rupture forces of the last unbinding event (pdf) were calculated as previously described. 31 The pdfs were fitted with the multiple Gaussian function. The loading rates were determined by multiplying the pulling velocity with the effective spring constant, ie, the mean slope at rupture.