Differences between the glycosylation patterns of haptoglobin isolated from skin scales and plasma of psoriatic patients

Abstract

Improved diagnosis of psoriasis, by new biomarkers, is required for evaluating the progression rate of the disease and the response to treatment. Haptoglobin (Hpt), a glycoprotein secreted by hepatocytes and other types of cells including keratinocytes, was found with glycan changes in psoriasis and other diseases. We previously reported that Hpt isolated from plasma of psoriatic patients is more fucosylated than Hpt of healthy subjects. The aim of this study was to compare the glycosylation pattern of Hpt isolated from skin scales or plasma of patients with psoriasis with that of Hpt from cornified epidermal layer or plasma of healthy subjects. High performance liquid chromatography analysis of the glycans isolated from the protein backbone revealed that glycan patterns from skin and plasma of patients were similar, and mostly displayed quantitative rather than qualitative differences from normal pattern. Biotin-labeled lectins were used to evaluate quantitative differences in the glycoforms of Hpt from plasma and psoriatic skin scales. Hpt from skin and plasma of patients showed more fucosylated and branched glycans than Hpt from plasma of healthy subjects. Tryptic glycopeptides of Hpt were also analyzed by mass spectrometry, and a decreased amount of sialylated glycan chains was found in glycopeptides of skin Hpt, as compared with Hpt from plasma. High levels of glycans with fucosylated and tetra-antennary chains were detected on the peptide NLFLNHSENATAK from Hpt of psoriatic patients. Our data demonstrate that specific changes in glycan structures of Hpt, such as enhanced glycan branching and fucose content, are associated with psoriasis, and that differences between circulating and skin Hpt do exist. A lower extent of glycan fucosylation and branching was found in Hpt from plasma of patients in disease remission. Altered glycoforms might reflect changes of Hpt function in the skin, and could be used as markers of the disease.

Figure 1. Normal phase HPLC pattern of 2AB-glycans from pHpt-N, pHpt-P and sHpt-P.

Purified pHpt-N, pHpt-P, and sHpt-P were deglycosylated by treatment with PNGase F, and their glycans were labelled by 2 aminobenzamide. After solid phase extraction, the glycans were fractionated by HPLC using a TSK-gel Amide-80 column (4.6×250 mm) with a linear gradient of ammonium formate at pH 4.4 (87.5 to 162.5 mM) with CH₃CN (65 to 35%) in 75 min, at 0.4 ml/min flow rate. Elution was monitored by measuring the label fluorescence at 425 nm (λ_EX = 360 nm). Panel A: glycans from pHpt-N. Panel B: glycans from pHpt-P. Panel C: glycans from sHpt-P. The GU ladder represents the migration of labelled oligosacharides with different units of glucose. The marked peaks were used for comparative analysis.

Figure 2. Binding of lectins to Hpt.

The wells of a microtiter plate were coated with different amounts of pHpt-N (white bars), pHpt-P (grey bars), or sHpt-P (black bars). Solutions containing 1 µM biotinylated LTA (panel A), MAA (panel B), ConA (panel C), or SNA (panel D) were separately incubated into the wells. Avidin-HRP and hydrogen peroxide were used to develop color from OPD. Color intensity was determined by measuring the absorbance at 492 nm (A₄₉₂). Five equal aliqouts of each sample were processed, and means ± SEM are shown. Asterisk: significant difference between the linked bars (P<0.0001). Triangle: not significant difference (P>0.05). Data from one experiment are shown. Inter-assay CV for each sample, from three separate experiments, was less than 5%.

Figure 3. Mass spectrum of the P1 glycopeptide repertoire from skin of patients.

Purified sHpt-P was digested by trypsin, and the resulting fragments were fractionated by UPLC and analyzed by ESI-MS. Positive ions of P1 (MVSHHNLTTGATLINEQWLLTTAK) glycopeptides from Hpt of skin of patients is shown. Glycopeptide peaks are indicated with their observed mass value (see Table 2) and chain structure. Peaks with mass attributable to non-glycosylated species were ignored.

Figure 4. Mass spectrum of the P2 glycopeptide repertoire from skin of patients.

Purified sHpt-P was digested by trypsin, and the resulting fragments were fractionated by UPLC and analyzed by ESI-MS. Positive ions of P2 (NLFLN HSENATAK) glycopeptides from Hpt of skin of patients is shown. Glycopeptide peaks are indicated with their observed mass value (see Table 3). The glycoforms found only in sHpt-P and not in pHpt-N and in pHpt-P are indicated. Peaks with mass attributable to non-glycosylated species were ignored.

Figure 5. Mass spectrum of the P3 glycopeptide repertoire from skin of patients.

Purified sHpt-P was digested by trypsin, and the resulting fragments were fractionated by UPLC and analyzed by ESI-MS. Positive ions of P3 (VVLHPNYSQVDIGLIK) glycopeptides from Hpt of skin of patients is shown. Glycopeptide peaks are indicated with their observed mass value (see Table 4) and chain structure. Peaks with mass attributable to non-glycosylated species were ignored.