Proc. Nati Acad. Sci. USAVol. 80, pp. 2752-2756, May 1983Medical Sciences

Carbonic anhydrase II deficiency-identified as the primary defectin the autosomal recessive syndrome of osteopetrosis with renaltubular acidosis and cerebral calcification

(erythrocytes/bone resorption/parathyroid hormone/isozymes/heterozygote detection)

WILLIAM S. SLY*, DAVID HEWETT-EMMETTt, MICHAEL P. WHYTEt, YA-SHIOU L. Yut, ANDRICHARD E. TASHIANt*Departments of Pediatrics, Medicine, and Genetics, Washington University School of Medicine, Division of Medical Genetics, St. Louis Children's Hospital, St.Louis, Missouri 63110; tDivision of Bone and Metabolism, Departments of Medicine and Pediatrics, Washington University School of Medicine and the JewishHospital of St. Louis, St. Louis, Missouri 63110; and tDepartment of Human Genetics, University of Michigan Medical School, Ann Arbor, Michigan 48109

Communicated by Donald C. Shreffler, January 24, 1983

ABSTRACT The clinical, radiological, and pathological find-ings in three siblings affected with the autosomal recessive syn-drome of osteopetrosis with renal tubular acidosis and cerebralcalcification have been reported. In an effort to explain the pleio-tropic effects of the mutation producing this disorder, we pos-tulated a defect in carbonic anhydrase H (CA II), the only one ofthe three soluble isozymes of carbonic anhydrase that is.known tobe synthesized in kidney and brain. We report here biochemicaland immunological evidence for the virtual absence of CA II inerythrocytes of patients affected with this condition, whereas CAI level is not reduced. Levels of CA II in erythrocyte hemolysatesfrom asymptomatic obligate heterozygotes are about half of nor-mal. These findings: (i) elucidate the basic defect in one form ofinherited osteopetrosis; (ii) provide genetic evidence implicatingCA II in osteoclast function and bone resorption; (iii) explain pre-vious observations that carbonic anhydrase inhibitors block thenormal parathyroid hormone-induced release of calcium from bone;(iv) clarify the role of renal CA H in urinary acidification and bi-carbonate reabsorption; and (v) suggest a method to identify het-erozygous carriers for the gene for this recessively inherited syn-drome.

Osteopetrosis is an inherited metabolic bone disease in whicha generalized accumulation of bone mass prevents normal de-velopment of marrow cavities and the enlargement of osseousforamena (1-3). It has been called "marble bone disease" be-cause the bones are very dense radiographically, although thebones typically have an increased susceptibility to fracture (2,3). Multiple genetic defects produce osteopetrosis but themechanism common to all the known forms of osteopetrosis isa failure of bone resorption (4). In man, two principal types ofosteopetrosis have been described. One is a dominantly in-herited, relatively benign condition which is often detected ra-diologically in asymptomatic adults (2, 3). A second type is therecessive, lethal, malignant form of osteopetrosis. In this form,osteopetrosis is usually present at birth, becomes symptomaticearly in infancy, and leads to death in infancy or early childhoodfrom infection or bleeding (3). Forms of osteopetrosis with clin-ical courses of intermediate severity have also been recognized(3). Various animal models of osteopetrosis have been identifiedincluding four different mutations that produce osteopetrosis inmice (4).

In 1972, three separate reports described a distinct form ofosteopetrosis that occurred in association with renal tubular aci-dosis (5-7). The pattern of inheritance was autosomal recessive.

The clinical course was not entirely benign, but the disease wascompatible with long survival and the hematologic abnormal-ities that dominate the clinical picture in the recessive lethalform of osteopetrosis were absent. One of these families wasnot described in a complete report until 1980 (8), by which timetwo of the three affected siblings had been found to have cal-cification of the basal ganglia. In the same year, Ohlsson et al.(9) independently reported three Saudi Arabian families in-volving first-cousin marriages that produced offspring with os-teopetrosis, renal tubular acidosis, and cerebral calcification, asyndrome for which they proposed the name "marble brain dis-ease.

In an effort to explain the pleiotropic effects of the mutationunderlying this disorder by a single enzyme defect, we pos-tulated a defect in, one of the three isozymes of carbonic an-hydrase (CA I, CA II, CA III) which are known to be under sep-arate genetic control in humans (10-14). This hypothesis seemedattractive for two reasons: (i) metabolic acidosis can be pro-duced by sulfonamide inhibitors of CA (12), and (ii) several re-ports have shown that CA inhibitors can block the parathyroidhormone-induced release of calcium from bone, suggesting arole for CA in bone resorption (15-17).

The relationship of CA deficiency to cerebral calcification wasless apparent, although it is known that CA II is present in brain(18) and that CA inhibitors inhibit cerebral spinal fluid pro-duction (19) and affect electrical activity of the brain (20). A de-fect in the CA II isozyme seemed most likely because this is themost widely distributed of the three known soluble isozymesof CA in human tissues (10, 11) and CA II is the only solubleisozyme so far identified in renal and brain tissue (18, 21, 22).In addition, a genetically determined, virtually complete ab-sence of CA I in mature erythrocytes has been found to haveno clinical consequences (23). Because both CA I and CA II areexpressed in human erythrocytes, it was possible to test thishypothesis by examining these isozymes in hemolysates of pe-ripheral blood from the family we reported previously (8).

In this report, we describe studies showing what appears tobe an almost complete absence of CA II in erythrocytes of pa-tients affected with the syndrome of osteopetrosis, renal tu-bular acidosis, and cerebral calcification. Furthermore, we re-port that in asymptomatic normal parents of affected patientsthe levels ofCA II are half of normal, which supports the inter-pretation that the CA II deficiency is the basic defect under-lying this clinical disorder and suggests a means to identify het-erozygote carriers.

Abbreviations: CA, carbonic anhydrase; PTH, parathyroid hormone.

2752

The publication costs ofthis article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

May

31,

202

1

Proc. Natl. Acad. Sci. USA 80 (1983) 2753

MATERIALS AND METHODSPreparation of Erythrocyte Hemolysates. Whole blood (about

10 ml) was collected in heparinized tubes (lithium heparin) andshipped by overnight carrier in ice. Immediately on arrival, theerythrocytes were washed three times in 2 vol of 0.85% NaClat room temperature. Hemolysates were prepared by lysing thecells with 1 vol of distilled water and extracting with 0.4 vol oftoluene. This mixture was shaken on a Vortex mixer for 1 minand then centrifuged at 2,250 X g for 20 min. After aspirationof the toluene layer and membrane interface, the aqueous layerwas centrifuged in a Microfuge at 12,000 X g for 5 min to re-move remaining cellular debris.

Electrophoresis, Staining, and Immunodiffusion. Verticalstarch gel electrophoresis of the hemolysates was carried out (8V/cm for 18 hr) with a borate buffer system at pH 8.6 (24). Theesterase activities of the electrophoretically separated CA I andCA II isozymes were detected with a mixture of 4-methylum-belliferyl acetate for CA I and fluorescein diacetate for CA II(25); the electrophoretograms were stained with 0.4% nigrosinto detect protein (24).The antisera to the CA isozymes were prepared as described

(26) by injecting rabbits with human CA I and CA II purifiedfrom hemolysates by affinity chromatography on sulfonamide-bound CM-Sephadex columns (27). Double immunodiffusionon agar plates (1.5% agar in 0.2 M sodium citrate-buffered sa-line, pH 6.7) was carried out by standard methods.

Quantitation of CA I and CA II. The ratio of CA I and CAII in individual hemolysates was determined by measuring theamounts of CA I and CA II after their separation by reverse-phase HPLC as described (28). Two hundred microliters of freshhemolysate was extracted with 200 ,ul of 40% ethanol (kept at-20°C) and 100 .ul chloroform (kept at -20°C) by vortexing ina polypropylene Microfuge tube for 1 min at 40C. A longer ex-traction time results in some loss of CA II (and a smaller lossof CA I); shorter extraction can result in a pink aqueous layerdue to retention of some hemoglobin components. Samples werecentrifuged at 12,000 x g for 2 min and the aqueous layer wascollected. Fifty-microliter samples were injected into an Altex-Beckman HPLC apparatus containing a Waters ,Bondapak C18column. Buffer A [0.1% trifluoroacetic acid (Pierce) in HPLC-grade water (Burdick and Jackson, Muskegon, MI)] and bufferB [0.05% trifluoroacetic-acid in UV-grade acetonitrile (Burdickand Jackson)] were used to develop, over 40 min, a 38-46%linear gradient of buffer B. The peaks were monitored at 215nm, and those corresponding to CA I and CA II were integratedby using a Hewlett-Packard 3390A integrator. The CA I/CA IIratios were calculated without any correction for the small (104, 9.0, 16.8, and 16.3,respectively. The virtual absence of CA II in the affected pa-tient was confirmed by this method. Because the higher-than-control ratios in the obligate heterozygote parents suggestedthat CA I/CA II ratios determined by this method might be

HbA---

HbA2 ---- CAI -

Origin

'

--CA If---

8 7 6 5 4 3 2 1

FIG. 2. Patterns of erythrocyte CA I and CA II of individuals shownin Fig. 1 and of controls, stained for esterase (A) and protein (B) afterstarch gel electrophoresis for 18 hr in 20 mMsodium borate pH 8.6 buff-er at 40C. Lanes: 1, unaffected control (spouse of m-3); 2, unaffectedfather (11-15); 3, unaffected mother (11-14); 4, affected proband (EI-1);5, unaffected sister (1-3); 6, affected sister (I11-5); 7, affected sister (III-6); 8, unaffected control (spouse ofm-6). See Figs. 1 and 5 for individualpedigree designations.

Medical Sciences: Sly et al.

Dow

nloa

ded

by g

uest

on

May

31,

202

1

Proc. Natl. Acad. Sci. USA 80 (1983)

A R

FIG. 3. Ouchterlony double-diffusion tests. Center wells: A, anti-serum against human erythrocyte CA I; B, antiserum against humanerythrocyte CA II. Outer wells in bothA andB contained hemolysatesfrom the following sources: 1, affected sister (E1-6); 2, unaffected control(spouse of III-3); 3, unaffected mother (11-14); 4, unaffected father (11-15); 5, unaffected control (spouse of III-6); 6, unaffected sister (II-3); 7,affected proband (III-1); 8, affected sister (III-5).

used to determine heterozygosity for the gene for this condi-tion, blood was obtained from additional members of this familyand CA I/CA II ratios were determined by HPLC. The ratiosfell into three groups: >104 for affected patients; 6.2-9.8 forone group which included four controls and five unaffectedmembers at risk for heterozygosity; and 13.4-17.2 for a groupwhich included two obligate heterozygotes and five unaffectedpeople at risk for heterozygosity (Fig. 5).

The mean value for the CA I/CA II ratio for controls ob-tained by using HPLC in this study is somewhat higher thanthat obtained by radioimmunoassay in a prior study [mean ±SEM: 6.32 ± 0.94 for 58 white control subjects and 6.64 ± 1.37for 79 black control subjects (24)]. This higher value may reflectthe preferential denaturation of CA II by the chloroform/ethanolextraction used to prepare hemolysates for the HPLC analysis.However, the absence of overlap in CA I and CA II ratios de-termined by HPLC between the small number of controls andthe presumed heterozygotes in this study suggests that the CAI/CA II ratio can be used to determine heterozygosity for thegene producing the inherited syndrome of osteopetrosis withrenal tubular acidosis and cerebral calcification.

DISCUSSIONAll three soluble isozymes of CA in humans, CA I, CA II, andCA III, are monomeric, =29,000-dalton, zinc metalloenzymeswhich catalyze the reversible hydration of CO2 (reaction I) (10-

1 2

I + +

E 1-13 1 51

16.8 16.3 7.8

2 3 4 5 6

Air>o10 1.7 6.4 >10' >10' 9

12 Cb (b114.5 13.4

300 0 15 30

O'4 i c.IAI 16.3012 44 012 44

000 ~~43 0910 - 43

006 42 0.06 42

I06 414.0-a 41

ol04 7 40 0.04 CAnl 40

6); 39ther3II15)9NtCAI i 9

[0 5 38 0-0 15 30 1~05 30TIME(minutes)

FIG. 4. Separation of CA II and CA I from hemoglobin-extractedhemolysates by reverse-phase HPLC. (Upper Left) Control (spouse ofmI-6); (UpperRight) mother (11-14); (LowerLeft) affected daughter (Ill-6); (LowerRight) father (11-15). Note the absence of CA II in the affecteddaughter. The CA I/CA 11 ratios were determined by integration of theHPLC peaks (see text). The value >104 for the affected daughter rep-resents a minimum obtained by using the smallest genuine peak in-tegrated (no peak was detected in the CA 11 region).

14). Reaction HI is an ionic dissociation that is virtually instan-taneous and is not subject to enzymatic acceleration.

I IICO2 + H20=±H2CO3 HCO3 +HC.

The direction of these reactions depends on the relative con-centrations of CO2 and HCO3 and on the pH. There is also adistinctive membrane-bound CA in lung, which .may representa fourth enzyme and tentatively is designated CA IV (29). The

HOMOZYGOUSCA II DEFICIENCY. OSTEOPETROSIS

[ O PRESUMED HETEROZYGOTE UNAFFECTEDD O PRESUMED NORMALS AND CONTROLS

FIG. 5. Extended pedigree of family with syndrome of osteopetrosis with renal tubular acidosis and cerebral calcification. Values beneath in-dividuals in the pedigree indicate CA I/CA II ratios determined by HPLC. (See legend to Fig. 4 and text for details of the HPLC analysis.) Familymembers II-14 and II-15 are obligate heterozygotes. The other heterozygotes indicated are presumed to be heterozygotes on the basis of similaritiesof their CA I/CA 11 ratios to those of obligate heterozygotes.

2754 Medical Sciences: Sly et al.

Dow

nloa

ded

by g

uest

on

May

31,

202

1

Proc. Natl. Acad. Sci. USA 80 (1983) 2755

kidney also contains a membrane-bound CA that is an intrinsiccomponent of the brush border of the proximal tubule (30-32).Genetic and structural evidence suggests that at least the sol-uble isozymes comprise a multilocus enzyme family derived froma common ancestral gene by gene duplications (11). However,the kinetic parameters of these isozymes and their sensitivityto inhibitors can differ markedly (33, 34). These findings havesuggested that the physiological roles for the different isozymesare diverse, which is also suggested by their different tissuedistributions.The human CA II isozyme, whose turnover number for the

CO2 hydration reaction under physiological conditions (1.3-1.9X 10' sec') is the highest known for any enzyme (35, 36), hasbeen identified (immunologically or by purification) in a widevariety of cells, tissues, and organs including erythrocytes, brain,eye, kidney, cartilage, liver, lung, skeletal muscle, pancreas,gastric mucosa, and anterior pituitary body (10, 11). The otherisozymes, whose activities toward CO2 and HCO3 are lowerthan those of CA II in the order CA II > CA IV > CA I> CAIII (29, 33, 34), appear to have a more limited distribution. CAI is found primarily in erythrocytes, CA III mainly in red skel-etal muscle, and CA IV in lung.

The finding of a quantitative defect in CA II in these patientsprovides us with an unusual opportunity to assess the impor-tance and function of this isozyme. In view of the high CO2 hy-drase activity of CA II and its wide tissue distribution, one mightexpect widespread effects of this deficiency in organs in whichCAII plays an important role. However, it should be noted thatthe finding of a virtual absence of CA II in erythrocytes doesnot necessarily imply a comparable deficiency in other tissuesand organs in which CA II has been reported. CA II levels incells with considerably more rapid turnover than erythrocytescould be appreciably higher. Also, there may be additional, stillunidentified, CA genes contributing to enzyme levels in tissuesspared by this mutation. However, what is clear from these pa-tients is that the quantitative deficiency of CA II that we havedemonstrated in erythrocytes has important clinical conse-quences for bone, for kidney, and for brain that merit discus-sion.Bone Metabolism. All known forms of osteopetrosis are as-

sociated with failure to resorb bone (4). Studies of several pa-tients with osteopetrosis have demonstrated impaired hyper-calcemic responses to infused parathyroid hormone (PTH) (37,38). The cause for this impairment might differ in the differentforms of osteopetrosis. Studies showing inhibition of PTH-in-duced release of calcium from bone by CA inhibitors have sug-gested a role for CA in bone resorption (15-17). Also, CA hasbeen demonstrated histochemically in chick and hen osteoclasts(39). On the basis of these and other observations (40, 41), ithas been suggested that PTH activates CA in certain bone cellswhere it might aid the resorptive process by mediating secre-tion of H' (16, 39). The genetic evidence presented here pro-vides strong support for a role of CA in bone resorption, whichwas suspected from the pharmacological and histochemical evi-dence cited above, and specifically implicates the CA II iso-zyme in bone resorption.

Renal Tubular Acidosis. There is general agreement that renalreabsorption of bicarbonate is a major factor in the maintenanceof acid-base homeostasis (42). Most of the bicarbonate recla-mation takes place in the proximal tubule and depends on CA(43). Only recently has it become clear how both a soluble (cy-tosolic) and a membrane-bound (luminal) CA play separate rolesin the proximal tubule (21, 42-46). Bicarbonate reclamation de-pends on H+ secretion, the major mechanism for proximal tu-bular acidification (42). The H' secreted into the lumen of theproximal tubule is titrated by the HCO3 in the glomerular fil-

trate to produce H2CO3 which is in contact with the membrane-bound CA. The luminal CA then catalyzes the dehydration ofH2CO3 to CO2 and H20 (42, 43). The CO2 diffuses freely intothe proximal tubular cell, where it can be hydrated by the cy-tosolic CA to H2CO3. Dissociation of this product into H' andHCO3 allows HCO3 to be transported by unknown mecha-nisms into interstitial fluid or the peritubular capillary, com-pleting the reclamation of filtered bicarbonate (42, 43). The re-generated H' can be secreted in exchange for Na+ to initiateanother cycle (42). From the above, it is clear how two differentCAs operate at different sites to participate in bicarbonate rec-lamation.The enzymatic dehydration of H2CO3 in the lumen appears

to be mediated entirely by the membrane-bound CA presentin the brush border of proximal tubular cells (42, 45, 46). Al-though we have no direct information on the status of thisenzyme in the patients described here, we have no reason tosuspect that this enzyme, which is biochemically and im-munologically distinct from CA II and the other soluble iso-zymes (21, 22, 31, 32), is defective in these patients. On theother hand, the enzymatic hydration of intracellular CO2 is pre-sumably mediated entirely by the soluble enzyme CA II, forwhich these patients are deficient. The renal tubular acidosispresent in patients with this syndrome must be explained inthis context.

Although the renal tubular acidosis in the different pedi-grees has been variable in severity, and somewhat heteroge-neous in type, most of the patients reported have a significantdistal defect (47). In those cases in which HCO3 reabsorptionhas been adequately studied, a proximal component, evidencedby HCO3 wasting at normal plasma HCO3 concentrations,has also been found (47). Thus, the patients appear to have amixed or hybrid type of renal tubular acidosis which includesboth a proximal component and a distal component. The hy-dration of CO2 in cells of the proximal tubule presumably ismediated by CA II which appears to be the major (and perhapsonly) soluble isozyme present in the kidney (21, 22, 48). Be-cause this reaction generates some of the H' secreted by theproximal tubule, and bicarbonate reclamation in the proximaltubule depends almost entirely on H' secretion (42-44), wecan understand why patients with CA II deficiency might havea renal tubular acidosis that includes a proximal component.The prominent distal component of the renal tubular acidosis

in CA II-deficient patients, evidenced by inappropriately highurine pH values when patients were quite acidotic (47), initiallywas more difficult to understand. A possible explanation wasprovided recently by immunohistochemical evidence showingmuch more intense reaction for CA II in the distal tubules (andeven in collecting ducts) than in proximal tubules of human kid-neys (48). These results suggest that CA II plays a more im-portant role in the distal tubule than was previously suspected(22), either in generating H+ or in titrating OH- produced bythe proton-translocating ATPase.The report (23) that a genetically determined, virtually com-

plete, absence of erythrocyte CA I in man is not associated withrenal tubular acidosis is consistent with biochemical and im-munological evidence that CA II is the only soluble isozymepresent in kidney (21, 22, 48). In view of this evidence, and ofthe findings presented here, it is difficult to understand the sig-nificance of the abnormalities in erythrocyte CA I that have beenreported in a few patients with distal type renal tubular acidosis(49, 50).

Brain Metabolism. The function of CA II in brain and thereasons for brain calcification in patients with defects in CA IIare less well understood. In the central nervous system, CA IIis primarily a glial enzyme and occurs predominantly in oh-

Medical Sciences: Sly et al.

Dow

nloa

ded

by g

uest

on

May

31,

202

1

2756 Medical Sciences: Sly et al.

godendrocytes (18). CA II has been identified in brain homog-enates, with up to 50% of the activity in a membrane-boundform (51). Although the functions of CA in brain are still spec-ulative, it is worth noting that many of the patients with thesyndrome of osteopetrosis with renal tubular acidosis and ce-rebral calcification have significant mental retardation (9). Thepatients in the family reported here are exceptional in this re-gard because their IQ scores are in the low normal range (8).Like the mechanism of the cerebral calcification in this syn-drome, the mechanism of the mental retardation is not yet clear.

Erythrocyte Function. One might expect some secondaryconsequences of CA II deficiency in tissues in which the CO2produced must be delivered to circulating erythrocytes and dis-charged from the lungs. This transport depends on the abilityof the CA activity in erythrocytes to convert metabolic CO2 toHCO3 rapidly in the tissues and to catalyze the reverse re-action in lung capillaries. Although CA II normally accounts foronly 14-17% of the CA in erythrocytes (CA I accounts for therest), it has been estimated that CA II accounts for about 90%of the CA activity of erythrocytes in vivo (36, 52). This estimateis based on the much greater specific activity ofCA II comparedto CA I and on the much greater sensitivity ofCA I to inhibitionby the normal chloride concentration of erythrocytes (52). Thisestimate is in general agreement with the findings by W. R.Chegwidden (personal communication), in a preliminary ex-periment on the family described here, that the relative CA ac-tivities for the HCO3 dehydration reaction in hemolysates froman affected homozygote (III-1), an obligate heterozygote (II-14),and an unrelated control were 0.17, 0.43, and 1.0, respectively,after correction for the nonenzymatic control reaction. Thus, allof the available evidence indicates that CA II is more importantthan CA I for the CO2 hydrase reaction in the erythrocyte.However, Wistrand (36) has estimated that only 2% of normallevels of CA activity in erythrocytes would be required for un-loading CO2 in lung capillaries at rest, and 4% would be re-quired at work. If all of these estimates are correct, CA I ac-tivity alone may be sufficient for this erythrocyte function. Thefact that the patients described here have no disability that wecan ascribe to the virtual absence of CA II in their erythrocytessupports this conclusion.We thank Patrick J. Venta and Dr. Alan Robson for helpful sugges-

tions, Mrs. Sabra Lovejoy for typing the manuscript, and Mrs. DorothyChesnut, RN, for obtaining blood samples. This research was supportedby National Institutes of Health Grants GM 21096, GM 31988 (to W.S. S.),and GM 24681 (to R.E.T.) and by the National Institutes of HealthClinical Research Center sponsored by Grant RR00036.1. Albers-Schonberg, H. (1904) Muench. Med. Wochenschr. 51, 365.2. Johnston, C. C., Jr., Lavy, N., Lord, T., Vellios, F., Merritt, A.

D. & Deiss, W. P., Jr. (1968) Medicine 47, 149-167.3. Beighton, P., Hamersma, H. & Cremin, B. J. (1979) S. Afr. Med.

J. 55, 659-665.4. Marks, S. C. (1982) Am. J. Anat. 163, 157-167.5. Guibaud, P., Larbre, F., Freycon, M. T. & Genoud, J. (1972) Arch.

Fr. Pediatr. 29, 269-286.6. Sly, W. S., Lang, R., Avioli, L., Haddad, J., Lubowitz, H. &

McAlister, W. (1972) Am. J. Hum. Genet. 24, 34 (abstr.).7. Vainsel, M., Fondu, P., Cadranel, S., Rocmans, C. & Gepts, W.

(1972) Acta Paediatr. Scand. 61, 429-434.8. Whyte, M. P., Murphy, W. A., Fallon, M. D., Sly, W. S., Tei-

telbaum, S. L., McAlister, W. H. & Avioli, L. V. (1980) Am. J.Med. 69, 64-74.

9. Ohlsson, A., Stark, G. & Sakati, N. (1980) Dev. Med. Child Neu-rol. 22, 72-96.

10. Tashian, R. E. (1977) in Isozymes: Current Topics in Biologicaland Medical Research, eds. Rattazzi, M. C., Scandalios, J. G. &Whitt, G. S. (Liss, New York), Vol. 2, pp. 21-62.

11. Tashian, R. E., Hewett-Emmett, D. & Goodman, M. (1983) inIsozymes: Current Topics in Biological and Medical Research, eds.

Rattazzi, M. C., Scandalios, J. G. & Whitt, G. S. (Liss, New York),Vol. 7, pp. 79-100.

12. Maren, T. H. (1967) Physiol. Rev. 47, 595-838.13. Lindskog, S., Henderson, L. E., Kannan, K. K., Liljas, A., Ny-

man, P. 0. & Strandberg, B. (1971) in The Enzymes, ed. Boyer,P. D. (Academic, New York), Vol. 5, pp. 587-665.

14. Pocker, Y. & Sarkanen, S. L. (1978) Adv. Enzymol. 47, 149-274.15. Waite, L. C., Volkert, W. A. & Kenny, A. D. (1970) Endocrinol-

ogy 87, 1129-1139.16. Waite, L. C. (1972) Endocrinology 91, 1160-1165.17. Minkin, C. & Jennings, J. M. (1972) Science 176, 1031-1033.18. Kumpulainen, T. & Nystrom, S. H. M. (1981) Brain Res. 220, 220-

225.19. Vogh, B. P. (1980) J. Pharmacol. Exp. Ther. 213, 321-331.20. Woodbury, D. M. & Kemp, J. W. (1970) Pharmakopsychiatr./

Neuropsychopharmakol. 3, 201-226.21. Wistrand, P. J. (1980) Acta Physiol. Scand. 109, 239-248.22. Dobyan, D. C. & Bulger, R. E. (1982) Am.J. Physiol. 243, F311-

F324.23. Kendall, A. G. & Tashian, R. E. (1977) Science 197, 471-472.24. Tashian, R. E. & Carter, N. D. (1976) in Advances in Human Ge-

netics, eds. Harris, H. & Hirschhorn, K. (Plenum, New York), Vol.7, pp. 1-56.

25. Hopkinson, D. A., Coppock, J. S., Muhlemann, N. F. & Ed-wards, Y. H. (1974) Ann. Hum. Genet. 38, 155-162.

26. Carter, N. D., Hewett-Emmett, D., Jeffery, S. & Tashian, R. E.(1981) FEBS Lett. 128, 114-118.

27. Osborne, W. R. A. & Tashian, R. E. (1975) Anal. Biochem. 64,297-303.

28. Hewett-Emmett, D. (1982) Fed. Proc. Fed. Am. Soc. Exp. Biol. 41,1385 (abstr.).

29. Whitney, P. L. & Briggle, T. V. (1982)J. Biol. Chem. 257, 12056-12059.

30. Sanyal, G., Pessah, N. I. & Maren, T. H. (1981) Biochim. Bio-phys. Acta 657, 128-137.

31. McKinley, D. N. & Whitney, P. L. (1976) Biochim. Biophys. Acta445, 780-790.

32. Wistrand, P. J. (1979) UpsalaJ. Med. Sci. Suppl. 26, 75 (abstr.).33. Koester, M., Pullan, L. M. & Noltman, E. A. (1979)Arch. Biochem.

Biophys. 211, 632-642.34. Sanyal, G., Swenson, E. R., Pessah, N. I. & Maren, T. H. (1982)

Mol. Pharmacol. 22, 211-220.35. Sanyal, G. & Maren, T. H. (1981) J. Biol. Chem. 256, 608-612.36. Wistrand, P. J. (1981) Acta Physiol. Scand. 113, 417-426.37. Aarskog, D., Aksnes, L., Haneberg, B. & Julshamn, K. (1979) Acta

Pediatr. Scand. Suppl. 277, 75-80.38. Caniggia, A., Gennari, C., Lore, F., Nuti, R. & Gall, M. (1980)

Eur. J. Clin. Invest. 10, 99-105.39. Gay, C. V. & Mueller, W. J. (1974) Science 83, 432-434.40. Dulce, H. J., Siegmund, P., Korber, E. & Schuitte, E. (1960)

Hoppe-Seyler's Z. Physiol. Chem. 320, 163-167.41. Forscher, B. K. & Cohn, D. V. (1963) in Mechanisms of Hard

Tissue Destruction (Am. Assoc. Adv. Sci., Washington, DC), pp.577-588.

42. DuBose, T. D., Jr., Pucacco, L. R. & Carter, N. W. (1981) Am.J. Physiol. 240, F138-F146.

43. Lucci, M., Pucacco, L. R., DuBose, T. D., Jr., Kokko, J. P. &Carter, N. W. (1980) Am. J. Physiol. 238, F372-F379.

44. DuBose, T. D., Jr., Pucacco, L. R., Seldin, D. W., Carter, N.W. & Kokko, J. P. (1979) Kidney Int. 15, 624-629.

45. Tinker, J. P., Coulson, R. & Weiner, I. M. (1981) J. Pharmacol.Exp. Ther. 218, 600-607.

46. Lucci, M., Tinker, J., Weiner, I., Crowther, C., Pulvino, J. &DuBose, T. D., Jr. (1982) Clin. Res. 30, 457 (abstr.).

47. Bourke, E., Delaney, V. B., Mosawi, M., Reavy, P. & Leston,M. (1981) Nephron 28, 268-272.

48. Spicer, S. S., Sens, M. A. & Tashian, R. E. (1982) J. Histochem.Cytochem. 30, 864-873.

49. Shapira, E., Ben-Yoseph, Y., Eyal, F. & Russell, A. (1974) J. Clin.Invest. 53, 59-63.

50. Kondo, T., Taniguchi, N., Taniguchi, K., Matsuda, I. & Murao,M. (1978) J. Clin. Invest. 62, 610-617.

51. Lees, M. B., Sapirstein, V. S., Reiss, D. S. & Kolodney, E. H.(1980) Neurology 30, 719-725.

52. Maren, T. H., Raybum, C. S. & Liddell, N. E. (1976) Science 191,469-472.

Proc. Natl. Acad. Sci. USA 80 (1983)

Dow

nloa

ded

by g

uest

on

May

31,

202

1