Human Adenosine Deaminase DISTRIBUTION AND PROPERTIES*

Adenosine deaminase exists in multiple molecular forms in human tissue. One form of the enzyme appears to be “particulate”. Three forms of the enzyme are soluble and interconvertible with apparent molecular weights of approximately 36,000, 114,000, and 298,000 (designated and respectively). The small form of adenosine deaminase is convertible to the large form only in the of a protein, which has an apparent molecular weight of 200,000 and has no adenosine deaminase This conversion of the small form of the enzyme to the large occurs at 4”, a pH optimum of 5.0 to 8.0, and is associated with a loss of conversion activity. The small form of the enzyme predominates in tissue preparations exhibiting the higher enzyme-specific activities and no detectable conversion activity. The large form of adenosine deaminase predominates in tissue extracts exhibiting the lower enzyme specific activities and abundant conversion The small form deaminase shows several electrophoretic variants by isoelectric focusing. The electrophoretic observed with the large form of the enzyme similar to that observed with the small form, the

EXPERIMENTAL PROCEDURE Materials-[B-"C]Adenosine (50.4 mCi/mmol) and [&"C]adenine (6.83 and 27.5 mCi/mmol) were purchased from New England Nuclear Corp. The [8-'dC]adenosine was diluted to a specific activity of 2 mCi/mmol with the nonradioactive compound before use. Amersham/ Searle Corp. provided    to . Adenine phosphoribosyltransferase activity and hypoxanthine-guanine phosphoribosyltransferase were assayed by a radiochemical technique (19). Cytochrome c oxidase was assayed according to the method of Cooperstein and Lazarow (20). Catalase activity was assayed at 25" by following the disappearance of hydrogen peroxide at 240 nm with a Zeiss spectrophotometer, as described by Beers and Sizer (21 A&i&y-Definition of factors involved in the interconversion of one molecular species of adenosine deaminase to another was achieved by subjecting preparations under specified conditions either to sucrose gradient ultracentrifugation or by applying an aliquot (50 ~1) of the preparation to one of several Sephadex G-100 columns (24 x 0.9 cm) layered with 1 cm of 13% agarose and equilibrated with 50 rnM Tris/HCl buffer (pH 7.4) containing 1 mM EDTA. Using column chromatography, the large molecular species, which eluted in the void volume (V,, 4.4 ml), could be separated from the small molecular form (V,, 9.4 ml)  Preparation-All steps were performed at 4" unless otherwise specified.
Non-neoplastic human tissue obtained post mortem was homogenized in 3 to 4 volumes (w/v) of Buffer A with 15 to 30 strokes of a Dounce homogenizer.
The homogenate was subsequently centrifuged at 6,600 x g for 20 min and the supernatant so obtained was stored at -70".
Storage at this temperature up to 6 months resulted in no loss of activity or alteration in the distribution of the molecular species of adenosine deaminase as determined either by gel filtration or sucrose density ultracentrifugation. After gel filtration, the fractions containing the activity of the appropriate molecular form of adenosine deaminase were pooled and concentrated approximately 5 to lo-fold by ultrafiltration as described above. Preparations were stable when stored at 4" or -70" for up to 6 months.
Conversion activity was prepared from human kidney tissue essentially by the method of Nishihara et al. (28). Frozen tissue was thawed and washed in 30 mM phosphate buffer (pH 7.4) (Buffer B). The tissue was then sliced and homogenized at 4" in 2 volumes (w/v) of Buffer B with a Sorvall blender at full speed for 1 min. The homogenate obtained was centrifuged at 30,000 x g for 30 min and the precipitate discarded.
The supernatant was brought to 60% saturation with ammonium sulfate and stirred for 2 h. The precipitate was removed by centrifugation at 30,000 x g for 20 min. Subsequently, ammonium sulfate was added to the supernatant to 80% saturation, stirred for 2 h and centrifuged as above. The precipitate was suspended in approximately 10 ml of Buffer B and subsequently dialyzed against 1000 volumes of Buffer A for 24 h. The dialyzed preparation was clarified by centrifugation at 30,000 x g for 20 min and concentrated P-fold by ultrafiltration as described above. The preparations were stable at 4" for up to 3 months.
Tissue Distribution-The appropriate tissues obtained at necropsy were sliced, washed twice with 154 mM sodium chloride, blotted dry with filter paper, homogenized at 4" in 3 to 4 volumes (w/v) of Buffer A with a Sorvall blender at full speed for 1 min, and centrifuged at 6,600 x g for 20 min. The supernatants obtained were dialyzed for 12 h against 1000 volumes of Buffer A and assayed for adenosine deaminase activity as described above. Subcellular Distribution-The activity and molecular forms of adenosine deaminase in the cytoplasm and subcellular organelles of human leukocytes was determined by differential centrifugation in isotonic sucrose employing a modification of the method of Morre (29).  determined by sucrose gradient ultracentrifugation of tissue extracts appeared to correlate with the specific activity of adenosine deaminase in the various tissue extracts (Table II). The large form of adenosine deaminase activity predominated in those tissue extracts exhibiting lower enzyme activity, e.g. lung or kidney, while the small molecular form was the major species in those tissue extracts exhibiting higher enzyme activity, e.g. stomach or spleen. Although these patterns of distribution were reproducible for most tissues, substantial variation was observed with extracts prepared from lung. In this tissue from different donors, either the large or small species could predominate despite attempts to maintain identical conditions. Subcellular Distribution-The subcellular distribution of adenosine deaminase in human leukocyte preparations is shown in Table III. The total recovery of adenosine deaminase activity in the various subcellular fractions was 58% of that present in the initial lysate prepared by freeze-thawing. This loss of activity probably represented inadequate disruption of the leukocytes since the recovery of protein and marker enzymes (cytosol: hypoxanthine-guanine phosphoribosyltransferase; plasma membranes: 5'-nucleotidase; mitochondria: cytochrome c oxidase) was comparable to the recovery of the adenosine deaminase activity. While the bulk of the total adenosine deaminase activity was present in the 100,000 x g supernatant, approximately 2% of the total activity present in the initial homogenate was associated with the 6000 x g and 100,000 x g pellets. This adenosine deaminase activity did not appear to be the result of contamination of these fractions with cytosol activity as no hypoxanthine-guanine phosphoribosyltransferase activity was detected in either fraction. Since the 6000 x g pellet exhibited substantial 5'.nucleotidase activity and attempts to reduce this apparent contamination have proven unsuccessful, the subcellular organelle(s) to which this form of adenosine deaminase is identified cannot be precisely defined. Sucrose gradient ultracentrifugation of the adenosine deaminase associated with the 100,000 x g supernatant revealed the presence of both the large and small forms but, in contrast to tissue extracts, there was no "particulate" adenosine deaminase activity sedimenting to the bottom of the gradient. An extract of the 100,000 x g pellet when subjected to sucrose gradient ultracentrifugation disclosed that the major Human Adenosine

Molecular
Deaminase 5451 and sedimented to the bottom of the gradient with a minor component exhibiting the sedimentation velocity of the small form. Gel filtration of splenic homogenate on 8% agarose revealed 5'-nucleotidase but not adenine phosphoribosyltransferase, hypoxanthine-guanine phosphoribosyltransferase, or cytochrome c oxidase in the void volume with a portion of the adenosine deaminase activity. Isolation of the adenosine deaminase activity eluting in the void volume followed by repeat gel filtration revealed, in addition to activity eluting in the void volume, the appearance of activity corresponding in elution volume to that of the small form (Fig. 3A). On other occasions, in addition to the small form, minor peaks of adenosine deaminase activity were discernable which corresponded in elution volume to the large and intermediate forms. Incubation of 'material appearing in the void volume with Triton X-100 followed by reIjeat gel filtration revealed as much as a 3-fold increase in the total adenosine deaminase activity eluting as the small form (Fig. 3, A and B). These observations suggest that the adenosine deaminase activity sedimenting to the bottom of the sucrose gradients or eluting in the void volume with gel filtration is associated with subcellular particulate matter.
Conuersion of Large to Small-The large molecular species of adenosine deaminase was isolated by gel filtration from extracts of tissues in which the small molecular form predominates. Following concentration by ultrafiltration, three peaks of adenosine deaminase activity were identified by either gel filtration on 8% agarose or sucrose gradient ultracentrifugation (Fig. 4). The sZO,w values and Stokes radii correspond to the large, intermediate and small forms of adenosine deaminase activity observed in the initial homogenate. When the small molecular species of adenosine deaminase newly formed from the large species was isolated by gel filtration and subjected to sucrose gradient centrifugation, a single peak of activity was observed with an s~,,,~ of 3.6. When the large form of adenosine deaminase (Stokes radius = 68, sza," = 10.4) was isolated by gel filtration of extracts prepared from lung or kidney tissue, in which this molecular species predominates, and then concentrated by ultrafiltration and subjected to further gel filtration or sucrose gradient ultracentrifugation, a single peak of activity was observed which had the Stokes radius and sZO.w of the large form. Unsuccessful attempts to dissociate this large molecular species obtained from kidney included incubation at 4" or 37" with increasing ionic strength (25 to 200 mM KC1 in Buffer A), the substrate or products of the reaction (2.5 to 10 mM adenosine, 5 to 10 mM inosine, 5 mM ammonium chloride), 5 mM 8-mercaptoethanol, or 5 mM p-chloromercuribenzoate. However, incubation of the large form of adenosine deaminase in 50 InM sodium succinate (pH 3.4) at 4" or 37" resulted in a decrease in the activity of the large form and the appearance of activity with an s 20,w of 3.6 ( Fig. 5) corresponding in sedimentation coefficient to that of the small form. The large molecular species was not dissociated after incubation under the same conditions in 50 mM sodium succinate (pH 4.4 to 5.0), 50 mM Mes (pH 5.5 to 6.5), or 50 mM Tris/HCl (pH 7.4 to 10.5). Repeat gel chromatography of "particulate" adenosine deaminase from human spleen with and without treatment with Triton X-100. The enzyme activity that eluted between 62 and 68 ml from the 8% agarose column (see Fig. 1 ing concentration by ultrafiltration, this form of the enzyme rechromatographed on 8% agarose as a single peak with a Stokes radius of 23 A. This same preparation with sucrose gradient ultracentrifugation yielded a single peak of adenosine deaminase activity with an szO+ of 3.6. On several occasions, a minor component (0.1% of the total activity) was identified which had an szO,w value of 10.4 and Stokes radius of 68 A. This small molecular form of adenosine deaminase was incubated for 60 min at 37" with tissue extracts in which the large molecular form predominates, such as kidney or lung (see Table I), and the reaction mixture subjected to gel chromatography. This led to a decrease in the peak of adenosine deaminase activity corresponding to that of the small form (Stokes radius 23 A) and the concomitant a pearance of peaks of activity with Stokes radii of 68 A and 38 K which correspond to the large and intermediate forms of the enzyme, respectively (Fig. 6). The large molecular species elaborated under these conditions exhibited an s 20,W of 10.7. This conversion of the small form to the large molecular species was independent of the tissue source of the small form of adenosine deaminase being demonstrable with the small form from human erythrocytes, spleen, stomach, or small intestine, as well as with the small molecular form elaborated upon the dissociation of the large molecular form under conditions described above. Incubation of the small molecular form of adenosine deaminase with the same concentration of either bovine serum albumin or cytochrome c, with increasing ionic strength, or with buffers over the pH range from 3.4 to 10.5, was ineffective in converting the small form to the large species. The failure to observe adenosine deaminase activity in the void volume of the 8% agarose column or at the bottom of sucrose gradients with ultracentrifugation when the isolated small, intermediate, or large forms were subjected to these procedures suggested that the "particulate" species did not represent a further aggregate of the soluble forms of adenosine The small form (1.7 kg, specific acti;ity 0.37 rmol/mi$&g of protein) isolated by gel filtration of a splenic preparation (see Fig. 1) was incubated at 37" for 60 min with 5 mg of bovine serum albumin in 10 mM Tris/HCl, pH 7.4 (A-A) or with 6.5 mg of human kidnev extract in 10 mM Tris/HCl, pH 7.4 (04). The samples were sequentially applied and eluted from the 8% agarose column equilibrated with Buffer A. The kidney extract (6.5 mg, O---O) was also incubated at 37" for 60 min and applied to the same 8% agarose column equilibrated with Buffer A. This was further supported by the absence of a "particulate" species following incubation of the small or large form with conversion activity.

Characteristics of the Conversion
Reaction-The distribution of conversion activity in different tissue extracts is shown in Table IV. Conversion activity was present in those tissue extracts in which the large species was the predominant form of adenosine deaminase, and was not detectable in those extracts in which the small form of adenosine deaminase predominated.
The partial purification of conversion activity (see "Experimental Procedure") eliminated more than 95% of the adenosine deaminase activity present in crude kidney extract and the resultant preparation exhibited a 2-to 3-fold increase in conversion specific activity with an overall recovery of 40%.
The time course of the conversion reaction at 4" and 37" and at two different concentrations of the preparation is shown in Fig.  7. The formation of the large form was complete within 5 min at 37" and within 15 min at 4'. Following completion of the reaction, the addition of increasing amounts of the small form led to no further formation of the large form. Under these conditions, the formation of the large molecular species was limited by the amount of protein present with conversion activity. However, by increasing the amount of conversion activity added to a constant activity of the small form, all of the small form could be converted to the large molecular form. Preincubation of conversion activity preparations with the large species of adenosine deaminase did not influence the amount of product formed on subsequent incubation with the small molecular species.
The effect of pH on conversion activity is shown in Fig. 8. The diminished conversion at pH 3.4 was not due to irreversible inactivation of either conversion activity or of the small form since incubation of the preparations at this pH in 50 mM sodium succinate for 20 min at 37" and subsequent dialysis against Buffer A resulted in complete recovery of both conversion activity and the adenosine deaminase activity of the small form. The conversion reaction did not require thiol compounds nor bivalent metal ions for activity. Preincubation with 5 mM p-chloromercuribenzoate, 5 mM dithiothreitol, or 5 mM p-mercaptoethanol and their subsequent removal by dialysis, did not affect the conversion activity; 5 mM EDTA, 2.5 mM MgCl*, or 2.5 mM CaCl, similarly were without effect. Various purine nucleosides and nucleotides including adenosine, inosine, AMP, IMP, GMP, cyclic 2':3'-AMP, ATP, and GTP at a final concentration of 2.5 mM had no effect on the conversion reaction.
Conversion activity present in kidney extracts was subjected to gel chromatography with 8% agarose on one occasion. The activity partitioned as a single peak which did not correspond to the peak of activity of the large molecular form of adenosine deaminase. The Stokes radius of conversion activity was 49 A. The conversion activity evident with gel filtration was pooled, concentrated by ultrafiltration, and subjected to sucrose gradient ultracentrifugation.
This yielded a single peak of conversion activity with an slO+ of 9.9. The calculated molecular weight and frictional ratio of conversion activity is listed in Table I.

Isoelectric
Focusing-Isoelectric focusing over a broad pH range (3.5 to 10) of crude tissue extracts exhibiting both the large and small molecular forms of adenosine deaminase disclosed all the enzyme activity between pH 4 and 6. Following isolation by gel chromatography, both the large and small molecular forms of adenosine deaminase also electrofocused within this pH range. Thus the remaining studies (with the one exception indicated) were carried out using ampholytes in the range from pH 4 to 6.
The small form of adenosine deaminase isolated from either small intestine or spleen exhibited two or three different electrophoretic forms (Table V). Sucrose gradient ultracentrifugation of each electrophoretic variant yielded a single peak of activity with an szO.w value ranging from 3.6 to 3.8, which corresponds to that of the native small form. Isoelectric focusing of the large molecular species of adenosine deaminase obtained by gel filtration from one of several different tissues revealed five to six different electrophoretic forms (Table V). The large form elaborated by incubation of the isolated small form with partially purified conversion activity yielded a pattern similar to that observed with the native large form. A quantity (1.1 units) of conversion activity partially purified from kidney tissue was incubated with 10 pg of a partially purified preparation of the small form (specific activity 6.3 pmol/min/mg of protein) for 20 min at 37" in a final volume of 0.1 ml containing 50 mM of the appropriate buffer (see Table VI) and bovine serum albumin (1 r&ml).
Sucrose gradient ultracentrifugation of each electrophoretic variant from kidney or the large form produced in vitro yielded a single peak of activity with an s~,,,~ value ranging from 10.6 to 10.8, which corresponds to that of the native large form. In contrast to the kidney gel filtration of the electrophoretic variants with p1 value(s) of 4.65 and 4.75 from lung indicated a mixture of the large and small molecular forms (ratio of activities being 0.64:0.36 and 0.37:0.63, respectively) while the electrophoretic variants with p1 value(s) of 5.06, 5.14, and 5.24 remained large. The results with large form from liver were essentially the same as observed with the large form from lung.
Kinetic Properties--Some of the kinetic constants of human adenosine deaminase are summarized in Table VI. The only  ' Adenosine deaminase was assayed at variable concentrations of [8-"Cladenosine ranging from 0.05 to 0.4 mM and fixed concentrations of inosine ranging from 1.0 to 5 mM. Protein concentrations used ranged from 6 to 30 pg.
b The variable buffers used at a final concentration of 50 mM were sodium succinate, pH 3.4 to 5.0; Mes, pH 5.5 to 6.5; Hepes, pH 7.0 to 8.0; Tris/HCl, pH 7.4 to 10.0.
distinct difference between the three soluble forms was the different pH optimum exhibited by the intermediate species.

DISCUSSION
The present study discloses that adenosine deaminase is widely distributed in human tissues with the highest activity evident in spleen and gastrointestinal tract. While the bulk of the activity is localized to the cytosol, up to 3% of the total cellular activity appears to be associated with a subcellular organelle(s).
Mustafa and Tewari (30) previously presented evidence for a mitochondrial form of adenosine deaminase in rat cerebral cortex, a finding questioned by others (31).
Four molecular species of human adenosine deaminase could be defined based on differences in molecular weight. One form appears to be "particulate" in nature based on the findings that (a) it has a molecular weight greater than 20,000,000, (6) it appears to be associated predominantly with a subcellular organelle, (c) treatment with a nonionic detergent is followed by the appearance of activity with a molecular weight of approximately 36,000, and (d) it could not be formed in reconstitution experiments. This form of the adenosine deaminase activity has not been recognized previously.
The remaining molecular species are soluble and have molecular weights estimated to be approximately 36,000, 114,000, and 298,000. The molecular weight of 36,000 for the In the present study we have found that the three soluble species of adenosine deaminase are interconvertible.
The large and intermediate forms of the enzyme dissociate spontaneously to the small form. Conversion of the small form of the enzyme to the large form appears to require another protein with a molecular weight of approximately 200,000. The interaction of conversion activity with the small form of adenosine deaminase does not appear to be a catalytic process since the reaction is only minimally temperature dependent and the conversion activity appears to be consumed in the process. Some of the characteristics of the conversion reaction reported here are similar to those described for the protein termed "conversion factor" purified from human lung (28). However, the data presented here suggests a molecular weight of 200,000 for conversion activity present in human kidney tissue, a value substantially different from the value of 139,000 for "conversion factor" described by Nishihara et al. (28).
In addition to species of different molecular weights, adenosine deaminase also exhibits substantial electrophoretic heterogeneity. Several forms of adenosine deaminase have been distinguished previously in erythrocytes by their electrophoretie mobility on starch gel electrophoresis (34). Tissues other than erythrocytes have been reported to exhibit additional forms termed "tissue-specific" isoenzymes which vary in their electrophoretic mobility in a manner specific for that particular tissue (13, 35). While considerable electrophoretic heterogeneity was demonstrated with each molecular form of adenosine deaminase using preparative isoelectric focusing, we were unable to confirm the presence of tissue-specific isoenzymes. The molecular basis for the electrophoretic heterogeneity of human adenosine deaminase also remains undefined.