High Resolution Gel Electrophoresis (HRE) on urine samples for qualitative analysis of proteinuria in dogs

Qualitative analysis of urinary proteins is mandatory to determine whether proteinuria depends on glomerular or tubular damage or on the presence of overflow proteinuria. To this aim, sodium dodecylsulphate gel electrophoresis (SDS) is considered the gold standard technique but it is expensive and timeconsuming. Aim of this study was to assess whether high resolution electrophoresis (HRE), a cheap and rapid technique, provides information about the presence or type of proteinuria. To this aim we measured Urinary Protein Creatinine ratio (UPC) and we ran both HRE and SDS-AGE on 87 samples of urine, 32 of which classified as non proteinuric (NP, UPC ratio <0.2), and 55 classified as either borderline proteinuric (BP, UPC ratio 0.2 to 0.5) or proteinuric (P, UPC ratio >0.5). Protein bands were detected by SDS in 14/32 NP samples (7 with glomerular, 3 with tubular, and 4 with mixed pattern) and in 52/55 P samples (14 with glomerular, 12 with tubular, and 24 with mixed patterns, and 2 with marked albuminuria). Using HRE, evident peaks were detected in 3/32 NP samples and in 40/55 P samples. Albumin, alfa1- and alfa2 globulins were most commonly detected, beta-globulin was rarely detected and only 1 sample contained gamma-globulin. The percentage of albumin in samples with glomerular pattern (51±17%) was significantly higher (P <.001) than in samples with tubular (32±24%) or mixed pattern (33±10%). By contrast, the percentage of alfa1-globulin in samples with glomerular pattern (24±12%) was significantly lower (P <.05) than in samples with tubular pattern (36±13%). No significant differences regarding alfa2-, beta- or gamma-globulin were found. The k coefficients of concordance between HRE or SDS and UPC were 0.59 and 0.55, respectively. Samples with detectable bands in SDS or in HRE but with UPC ratio <0.2 were classified as ‘‘false positive’’ in respect to the classification based on UPC ratio and samples without detectable bands but with UPC ratio >0.2 as ‘‘false negative’’. Using these values, sensitivity, specificity, and positive likelihood ratio were calculated and accounted for 73%, 90%, and 7.8 for HRE and 94%, 56%, and 2.2 for SDS. These results suggest that HRE and SDS have a comparable concordance with UPC ratio in detecting proteins in samples of canine urine. SDS, however, evidences protein bands also in samples that are actually NP. SDS should thus be performed only when UPC ratio is greater than 0.2, to provide additional information about the type of proteinuria. Due to its higher specificity, HRE can be used as a preliminary screening tool in samples with low pre-test probability of proteinuria, but its low sensitivity risks missing positive bands in samples that are actually proteinuric. These results encourage further studies on the concordance between HRE and the actual type or degree of renal disease and/or on the use of HRE in monitoring proteinuric patients.