The Purification and Characterization of Bovine Enterokinase from Membrane Fragments in the Duodenal Mucosal Fluid*

Bovine enterokinase has been purified from the mu- cosal fluid adhering to the intestinal wall. Enterokinase is predominantly present as membrane fragments which must be treated with Triton X-100 to release the enzyme. The purification resulted in a higher yield of enzyme in fewer steps and in less time than when mucosal cells were used. The properties of the enzyme in the fluid are identical with those found previously with the mucosal cell preparation but differ in the size of the subunits and in amino acid composition from the enzyme purified from intestinal contents It is highly unlikely that the existence of isoenzymes could explain these differences. It is more likely that the enzyme isolated from the intestinal contents represents an ex-tensively degraded form with retention of enzymatic activity. Enterokinase (enteropeptidase, activates pancreatic trypsinogen and initiates the intestinal digestion of proteins (1, 2). The enzyme is found in the duodenal brush-border membrane and the intestinal contents. Kuntiz (3) was the first to partially purify the porcine enzyme from the intestinal contents. Maroux et al. (2) developed a procedure for the preparation of a highly purified sample and also isolated the enzyme from a detergent extract of the duodenal mucosa (5). The mucosa was chosen as the source for the purification because it contained greater amounts of the enzyme. More recently, the bovine enzyme was purified from intestinal contents and from mucosal

creatic trypsinogen and initiates the intestinal digestion of proteins (1, 2). The enzyme is found in the duodenal brushborder membrane and the intestinal contents. Kuntiz (3) was the first to partially purify the porcine enzyme from the intestinal contents. Maroux et al.
(2) developed a procedure for the preparation of a highly purified sample and also isolated the enzyme from a detergent extract of the duodenal mucosa ( 5 ) . The mucosa was chosen as the source for the purification because it contained greater amounts of the enzyme.
More recently, the bovine enzyme was purified from intestinal contents (6) and from mucosal cells (7). Although the two preparations had many properties in common, they showed several significant differnces.
The mucosal and the intestinal content enzymes differed in the size of the heavy chain (115,000 and 82,000 daltons) and light chain (35,000 and 57,000 daltons). The amino acid compositions also differed.
In an attempt to better understand the relationship between the two forms of the bovine enzyme, we considered the possibility that isoenzymes exist in the cow (7). In our earlier studies, we examined all fractions for enterokinase activity in the course of purification from mucosa, but failed to find a second form of the enzyme. In the hope that we could resolve the differences, we now have examined the properties of the enzyme in the viscous fluid adhering to the lining of the intestinal wall. We selected this source because it contains the enzyme soon after release from the microvilli and reasoned that it could contain a form of the enzyme that is not membrane-bound and that could differ from the mucosal cell preparation. These studies of multiple forms of the enzyme are important in light of the species differences that have been found. For example, the molecular weight of the enzyme from the cow is approximately 150,000 (6, 7 ) , the pig 195,000 (5), and the human 300,000 (8). One would have expected greater similarities since the enzyme performs the same physiobgical function in these species. The studies described below show that the enterokinase molecule from the mucosal fluid is identical with that found previously from the mucosal cells but different from that in the intestinal contents.

EXPERIMENTAL PROCEDURES AND
RESULTS'

DISCUSSION
The approach taken in these studies was to characterize enterokinase from the intestinal mucosal fluid to see if the enzyme is identical with or related to the preparation obtained from the intestinal mucosal cell (7). The properties of the purified mucosal fluid enterokinase were identical with the mucosal cell enzyme with respect to the molecular weights of the intact enzyme, as well as the heavy and light polypeptide chains, the amino acid composition, and the enzymatic activity toward trypsinogen and several synthetic substrates. Both were inhibited to the same extent when titrated with pancreatic trypsin inhibitor and an active site titrant.
The identical size of the heavy and light chains of the two enzyme preprations eliminates a structural modification on release from the brush-border membrane. It is known that the enzyme is released from brush-border membranes by limited proteolysis with a loss of the membrane anchor (18, [39][40][41]. The mucosal fluid enterokinase has a polypeptide anchor, since the enzyme is incorporated into reconstituted phospholipid vesicles. We concluded that the desquamation of the tips of the brush-border membrane microvilli releases the intact enzyme as membrane fragments into the mucosal fluid. Only later, when the enterokinase enters the luminal contents, will the phospholipid-enzyme complex dissociate, releasing the free enzyme. The identical properties of the mucosal fluid and the mucosal cell enterokinase decrease the possibility of finding an isoenzyme. It seems reasonable that the enzyme isolated by Anderson et al. (6) from the intestinal contents represents a molecule altered by limited proteolysis and/or by degradation of the polysaccharide chains. Proteolysis in the intestinal contents is highly likely since the digestive enzymes are present in large amounts. Various enzymes that degrade carbohydrates are also present. Since both the heavy and light chains of enterokinase contain approximately 35% carbohydrate, both chains could be modified.
Finally, the activation peptide sequence of trypsinogen is highly conserved in different species. The sequences of Asp,-Lys found in most species or the Gluz-Asp-Lys and Glu-Asp2-Lys of lungfish trypsinogen all have a cluster of acidic residues preceding the lysine (42)(43)(44). The subsite specificity of enterokinase must recognize these positions (2, 45), because the enzyme has a high catalytic efficiency toward these sequences relative to the low efficiency of bovine trypsin. Thus, the structure of enterokinase, at least in the region of the substrate binding and active site residues, must also be conserved and a physiological function for isoenzymes does not appear likely. One hundred flfty nl of the formate buffer were collected In e beaker contarnlnq 37 ml of 0.5 n TTLJ, pH 9.0.

THE PURIFICiiTION AND CHARRCTERIZITION OF BOVINE ENTEROKINASE
The fraction was dlalyzed dqlln5L water. concentrated to 100 m l . fiiLered through a 2 um PolyYic nllllpore fllrer, and concentrated Lo less than 10 ml before lyophlliratlon.
was used for the purlflcatlon of the enzyme.
The method of Liepnieks and llght ( 7 1 was followed when deoxycholate suggested from the ldentlflcafian of brush border membrane enzymes l n the fluid. The total Units and specific actlvlty of enteroklnase and alkallne pnosphardee ~n the flvxd were compared to those present ~n the mucosal calls amaunts and with a hlgher specifac acrlvity. The detectlo" of membrane (Table 1) The hlqher yield Of enryne u s~n g the fluid Was due prlmarlly to the increased amount of enzyme ~n the startlng materlal.
T11Lon X-100 has also been used to solubillze other lntestlnal brush border membrane prOLelnJ 128. 29.301 and to partially p u r~f y enteraklnase ( 3 1 1 . The use of TrltOn X-100 eliminated the need to remove deoxycholate by acld precipltatlon which alvayl caused a small loss of enzyme actlvlty I l l . The ne* Source of the enzyme had the deczded advantage of beqlnnlng the 1 5 0 1 a t~o n procedure vlth a hlqher specif~c activ~ty of 2 3 3 un~rs/mq, compared vlth the 9 unrts/mg I" the extract of the mucosaI cells. The active fractron from the anLon exchange res," was also rncreased to 2910 un~re/rng compared Wlth the 1110 U n l t l / m g for the mucosal cell preparation. The affln~iy chromatography step produced a homogeneous enzyme of increased Speclflc actlvlty 1324.000 Y s 280,000 anltS,ng! and yleld I 3 6 V Q 2 8 % ) .