-I In normal dogs and in some dogs with marked hyperamylasemia there is very little urinary amylase a c t i ~ i t y . ' . ~ . ~ . ~ . ~ ~ This is surprising since the molecular weight of purified canine pancreatic amylase determined using a denaturing technique is about 54,000 daltons) which is below the threshold of the glomerular filter. It has been hypothesized that canine amylase does pass the glomerular filter but is totally reabsorbed from the filtrate or is ina~tivated.~ Recently, it has been shown that these mechanisms probably do not play an important role in amylase clearance and that canine amylase probably does not pass the intact glomerular filter.6 It appears that unique characteristics of the canine glomerular filter and amylase might account for the lack of appearance of urinary amylase activity. It has been reported that native human amylase appears as a polymer.8 Since the polymerization phenomenon may be important in understanding the clearance of canine amylase, an estimate of the molecular weight of canine serum amylase and the influence of this weight on glomerular filtration were determined. Four pools of canine sera, each composed of ten canine sera chosen randomly from routine laboratory accessions at the University of Guelph Veterinary Teaching Hospital, were made. Random urine samples were obtained from 13 normoamylasemic 1-year-old Samoyed dogs in a colony bred to propagate the sex-linked trait for hereditary glomerulopathy.' Urine samples were obtained from 16 normoamylasemic, non-hematuric, non-hemoglobinuric dogs with proteinuria. Total amylase activities were determined using a p-nitrophenylmaltohexaoside substrate (Coulter Electronics of Canada, Ltd., Burlington, Ontario, Canada). Urinary protein concentrations were determined by a Ponceau S (Aldrich Chemical Co., Milwaukee, WI) dye binding method and purified human albumin standards (Dade Diagnostics, Inc., Aguada, Puerto R i ~ o ) . ~ Urinary specific gravities were determined refractometrically. The four pools of canine sera were each diluted 300 times in a buffer containing 0.086 M NaC1, 7 mM KCl, 2 mM MgCl,, 7 mM CaCI,, and 0.02 M tris(hydroxymethy1)-aminomethane-HC1 (pH 7.2). Aliquots of diluted pooled sera were concentrated to the original volume of serum using 100 and 300 kilodalton molecular weight cut-off ultrafilters (Amicon Canada Ltd., Oakville, Ontario, Canada). Prior to use, 10 ml of a 0.15 M NaCl solution containing l.Oo/o bovine serum albumin was passed through each ultrafilter to prevent non-specific adherence. The ultrafiltration was done at 4 C and 10 Ib/sq. in. About 20% of albumin is retained by both sizes of ultrafilters (Amicon Canada Ltd., Oakville, Ontario, Canada). Amylase activities in the retentates were determined and results expressed as the percent amylase activity retained by each filter. The relations between urinary protein concentration, specific gravity, and amylase activity were determined by linear --