Background and Aims
Being a well-known intrinsic property of amyloid, сongophilia was recently demonstrated in urine of patients with proteinuria of different etiology including renal amyloidosis (RA) and non-amyloid nephropathies (NANP) [1]. Urine proteins (UPs) responsible for congophilia in RA and NANP are supposed to have another property of amyloid such as resistance to ionic detergents and present in urine as detergent-resistant aggregates (DRA). In the pilot study we performed mass-spectrometry (MS) analysis of congophilic urine samples and its detergent-resistant fraction in patients with RA and NANP to investigate UPs and their specificity to the particular kidney disease.

Method
We collected first morning void urine samples from patients with RA (n = 4) and NANP (n = 4). Urine congophilia was assessed by Congo red Dot test as described previously [1]. To analyze bulk urine (BU) proteins, 30 μl of supernatant was precipitated with 80% acetone and boiled in 2% sodium dodecyl sulfate (SDS) for 30 minutes. Then 5 μg of UPs was separated in 10% polyacrylamide gel electrophoresis (PAGE) followed by Coomassie blue staining. The ultracentrifugation of 0.5-3 ml of urine at 300,000x g for 16h following by the treatment of precipitate with 3% sarcosyl in phosphate-buffered saline (PBS) for 10 min with subsequent washing of resistant aggregates in PBS were applied to obtain DRA. After boiling in SDS 8 μl of the sample was separated in 10% PAGE. UPs concentration in the DRA was estimated by densitometry regarding to 2 μg of bovine serum albumin. Then 6 μg of UPs was digested with trypsin, followed by purification on silicate (CDS Empore™ C18 Extraction Disks). Prepared samples of BU and DRA were analyzed by electrospray ionization tandem MS. We used MsFragger software to obtain lists of UPs for each sample. Most representative UPs in the sample were selected by the unique spectral count (USC). We considered the protein as having diagnosis-specific potential (DSP) if it was present in every sample of the particular disease group, i.e. AA, AL, NANP in BU or DRA, and was absent in any sample of another disease group.

Results
The patients had following characteristics: age 51±13 years, 3 male/ 5 female, eGFR = 45 (19; 95) ml/min/1.73 m2; 24 h proteinuria = 6.5 (4.5; 7.9) g. In PAGE analysis, BU proteins appeared to be similar in all 8 samples with 2 major bends in 70 kDa and 50 kDa regions, corresponding to albumin and immunoglobulin (Ig) heavy chains, respectively (Fig. 1A). The amount of protein in DRA was small and comprised 1.02 (0.71; 1.61) % of the total protein in the sample (4.7 (2.3; 5,1) g/l). Compared with BU PAGE analysis of DRA proteins revealed other bends with trace albumin bend, predominance of 45 kDA region bend and more apparent 30 kDa bend (Ig light chains) in the majority of patients (Fig. 1B). Results of MS analysis are shown on the Figure 2. There were more DSP revealed in DRA vs BU: 14 vs 5, 46 vs 2 and 4 vs 2 UPs in samples of AL, AA and NANP, respectively. When compare by a particular disease group, there were no DSP found either in BU or DRA in RA samples. One protein (aminopeptidase N) and 3 proteins (isoform 3 of unconventional myosin-LC, protein S100-A8, elongation factor 1-α 1) were detected as DSP in BU and DRA in AL, respectively. Alpha-1-acid glycoprotein 1 was only DSP for AA in DRA as well as serum paraoxonase 1 for NANP in BU. Venn-diagrams of shared and divergent UPs are present on the Figures 2D-G.