Clinical characteristics of patients

All patients diagnosed with CP [n = 2824, age 35.6 ± 13.8 years; male 2024 (72.2%)] were screened for the presence of calcified/soft stones by imaging modalities such as ERCP and endoscopic ultrasonography (Fig. 1 A,B,C). The clinical parameters, along with their significance for CP patients without stones, with protein plugs and calcified chronic pancreatitis are provided in Table 1. A total of 1924 patients (68%) had calcifications/soft stones and 900 (32%) CP patients did not have calcifications/soft stones. Out of 1924, 1678 (59.0%) patients had calcified stones, and 246 patients (41.0%) had both soft stones and calcified stones. The prevalence of soft stones/protein plugs was higher in males (n = 192, 78%; 35.9 ± 13.8 years) compared to females (n = 54, 22%; 27.9 ± 13.5 years): P ≤ 0.0001. Lymphocytes were significantly high in CP patients with protein plugs and calcified stones, while neutrophils were high in CP patients without any stones. HbA1c values above 5.7, indicating a pre-diabetic (PD) status, and above 6.4, indicating a diabetic (D) status, were more common in CP patients in which all stones were calcified (PD: 23%, D: 57%) compared to CP patients with soft and calcified stones (PD: 22%, D: 54%) (Fig. 1D, E).

Table 1 Clinical parameters of CP patients without stones, CP patients with protein plugs and patients with calcified chronic pancreatitisFig. 1

ERCP images: A protein plug, B radiolucent stone, C calcified stone and HbA1c status: D box and whisker plot of HbA1c values of CP patients with both soft and calcified stones. (E) Box and whisker plot of HbA1c values of CP patients with calcified stones

Gel electrophoresis analysis of isolated proteins by PAGE and 2-dimensional (2D)

Pooled analysis of proteins isolated from protein plugs/soft stones showed 2 prominent bands at 14 and 26 kDa regions (Fig. 2A). These samples were then used for 2-dimensional electrophoresis to obtain a better resolution for the protein mixture. The 2D gel (Fig. 2C) showed seven prominent and highly intense spots with similar molecular weights (10.4 kDa) at different PIs ranging from 4 to 7. Five prominent and intense spots were seen at < 10 kDa with PIs ranging between 4.5 and 5.5. Eight tiny but intense spots were seen at 29 kDa with PI ranging between 4 and 7. Three 3 spots were observed at 27 kDa with PI ranging between 5.5 and 6.5. A total of 23 spots with high intensities were chosen for in-gel digestion and identification by MALDI-TOF. Low-intensity spots were excluded because the quantity was not sufficient for MALDI-TOF.

Fig. 2

Protein profile: A protein plugs and B pancreatic juice on a 10% SDS PAGE gel; M is the protein marker. 2-Dimensional gel image of C soft stone/protein plugs and D pancreatic juice

MALDI-TOF of protein plugs

Mass spectra of peptides obtained after trypsin digestion of twenty-three gel pieces were used to generate the protein mass fingerprint (PMF). The sequences of 14 out of 23 spots of protein plugs matched with Regenerating family member 1 alpha (Reg1A) (60.8%), which is also known as lithostathine. The sequences of all the lithostathine spots had an initial starting position from the 34th amino acid, indicating that the protein cleaved during the process of precipitation and plug formation. Tryptic digestion during in-gel digestion yielded 6 to 11 peptide fragments from the respective 2D gel spots with molecular weights of < 10, 10.4, 27 and 29 kDa and at PIs ranging between 4.17 and 6.97. The Mascot® scores of the identified proteins were between 100 and 170. The sequences of the protein fragments with score greater than 100 is presented in Table 2.

Table 2 Peptide mass fingerprint (PMF)Gel electrophoresis analysis of pancreatic juice protein

To clarify the protein profile of pancreatic juice, the isolated proteins were assessed by SDS PAGE and 2D gel electrophoresis. Pancreatic juice from patients without stones showed 6 prominent bands at 71 kDa, 55 kDa, 30 kDa, 32 kDa, 23 kDa and 14 kDa (Fig. 2B). The proteins isolated from pancreatic juice were subjected to 2D gel electrophoresis with pH values ranging from 4 to 7 to assess the protein profile (Fig. 2D). Nine spots of proteins were seen at < 10.4 kDa with a PI range of 5–6.5. Fifteen spots of proteins were seen between 20 and 40 kDa with a PI range of 4–7. Five spots of proteins were at the 51 kDa position with a PI range of 4.5–6.5. A total of 29 spots with high intensities were excised, used for in-gel digestion and eventually identified by MALDI-TOF.

MALDI-TOF of pancreatic juice

Twenty-nine in-Gel tryptic digested pancreatic juice spots were analysed using MALDI-MS. The resultant mass spectral data (PMF, Additional file 1: Fig. S6) were searched against the SWISS-PROT database using the MASCOT search engine. Of the twenty nine gel spots, eleven spots were identified as amylase, carboxypeptidase, trypsin, chymotrypsin and chymotrypsinogen. Out of the eleven identified spots, only one spot was identified as lithostathine (Reg1A).

In vitro precipitation of pancreatic juice

Protein precipitation was greater at trypsin concentrations between 8 and 12 µg/mL and incubation for 30 min. Therefore, the pancreatic juice was treated with 10 µg/mL trypsin, and its pH was varied from 4 to 8. After electrophoresis, protein degradation of protein was observed at pH 4, and few bands were visible at pH 5–8 (Additional file 1: Fig. S3). A single band was exclusively seen at pH 6 and 7 at the 18 kDa position. The protein was identified through excision of this band, gel digestion and MALDI-TOF of the extracted peptides (Fig. 3A). Four peptide fragments matched the sequence of chymotrypsin C (CTRC) with a molecular weight of 18 kDa and a score of 61. The sequences that matched CTRC were 113–119 (WNALLL. N), 152–162 (DYPCYVTGWGR. L) 173–188 (LQQGLQPVVDHATCSR. I), 189–195 (IDWWGFR.V).

Fig. 3

A Positive-mode MALDI-reflectron TOF mass spectra using HCCA as a matrix (1:1, v/v) of differentially expressed proteins observed at pH 6 and 7. B Hypothetical model of lithostathine precipitation and plug formation within the ductal lumen in CP patients

In silico analysis

Docking truncated lithostathine, which was identified in soft stones (protein plugs) and the native lithostathine (seen in pancreatic juice), with calcite molecule predicted a higher binding affinity for truncated lithostathine compared that of its native form. The binding energy of truncated lithostathine was − 3.8, whereas that of native lithostathine was − 2.9. β-D- Galactopyranose (NDG), sialic acid (SIA) and glutamic acid (Glu) at the 1, 3 and 6 positions were observed to interact with the calcite molecule of native lithostathine protein, whereas asparagine (Asn), leucine (Leu), 2 molecules of serine (Ser) and lysine (Lys) at the 49, 50, 52, 93, 141 positions, respectively, were predicted to interact with the truncated lithostathine protein. The interaction of both forms of lithostathine with calcite molecules is depicted in Fig. 4.

Fig. 4

In silico analysis: Docking of native (A and C) and truncated (B and D) lithostathine with calcite molecules (green and red triangular molecules) present in the right corner of image C and in the centre of image D, showing the interaction of calcite molecules at different positions on lithostathine using AutoDock Vina and PyMOL software