Eur. J. Biochem. 74, 603-610 (1977)

The Purification and Properties of Human a1-Antitrypsin (a1-Antiprotease), Variant Z Albert HERCZ and Marcela BARTON Research Institute, The Hospital for Sick Children, Toronto, and Department of Clinical Biochemistry, University of Toronto (Received July 15/0ctober 27, 1976)

After three stages of preliminary purification, variant Z was chromatographed on a DEAEcellulose column. Upon elution with a linearly increasing concentration of NaCI, variant Z was recovered in two separate peaks, the first of which contained 81 % and the second 19 % of the total. The preparation corresponding to the first peak was homogeneous by various criteria. The trypsin and chymotrypsin inhibiting capacities and the specific antigenic activity of the preparation were nearly the same as those of an authentic sample of variant M. Variant Z contained 8 or 9 more glycine residues than variant M, but no appreciable difference was found between their carbohydrate contents. By analytical isoelectrofocusing the isoinhibitors of purified variant Z overlapped with those in the plasma of the donor and were cathodal to, but partially overlapped with purified variant M. After desialysation, the overlap between the different variants became complete, but variant Z contained a larger proportion of cathodal and smaller proportion of anodal components than variant M. Both variants formed five distinct isoinhibitor-protease complexes after incubation with trypsin and chymotrypsin and the corresponding complexes in the different variants completely coincided. Over 20 different phenotypes of human al-antitrypsin can be distinguished by different electrophoretic and isoelectrofocusing methods (discussed in [I]). Some of the alleles of al-antitrypsin leading to a decreased plasma level of this protein, namely Piz (Pi standing for proteinase inhibitor) [2 - 41, Pis [2 - 41, Pio [5-91, PiM Malton [lo] and PiM Duarte [ I l l attract special attention because of their medical as well was their theoretical implications. An outstanding discovery of the past few years in this field was the recognition of massive intrahepatic accumulation of al-antitrypsin in carriers of the Piz allele [12- 151. The eel-antitrypsin in the serum of ZZ homozygotes is characterized by a slower (cathodal) than normal electrophoretic mobility [16- 191. Several authors attempted to bring these observations to a common denominator by suggesting that both would arise from impaired sialysation of variant Z [20-221. On the basis of the fact that desialysed glycoproteins bind abnormally with liver plasma membranes [22- 241, they proposed that a decreased sialic acid content of variant Z would lead to the intrahepatic accumulation, Enzymes. Trypsin (bovine pancreas) (EC 3.4.21.4) ; a-chymotrypsin (bovine pancreas) (EC 3.4.21 . l ) ; neuraminidase (Vibrio commu cholerae) (EC 3.2.1.18).

decreased plasma level and decreased (cathodal) electrophoretic mobility of this protein [20,21]. This suggestion gained further credence by the observation that sialyltransferase activities in some scl-antitrypsindeficient members of a family were appreciably decreased [25]. While this suggestion has had attractive features, on the whole it turned out to be untenable. As pointed out lately [26,27] the liver inclusion bodies in xlantitrypsin-deficient individuals are localized in the rough endoplasmic reticulum but have never been seen in the Golgi system of the hepatocyte, hence any possible alteration in the carbohydrate composition of this glycoprotein would be secondary to a primary defect of the peptide chain biosynthesis. In this work we purified variant Z from the plasma of a ZZ phenotype donor and compared it in various ways with purified variant M to gain insight into the nature of the defect leading to a1-antitrypsin deficiency. MATERIALS AND METHODS Materials

Purified bovine trypsin, 192 U/mg (9 JB) and x chymotrypsin, 61 U/mg (34 J 898) were purchased

604

from Worthington Biochemicals. Neuraminidase (mucopolysaccharide N-acetyl neuraminyl hydrolase) free of protease, aldolase and lecithinase C activity, prepared from Vibrio comma cholerae, with an activity of 500 U/ml was supplied by Behringwerke (Marburg). N-Acetyl-L-tyrosine ethylester was obtained from Sigma Chemicals and p-tosyl-L-arginine methylester from Schwarz and Mann. Acrylamide and N,N-methylenbisacrylamide used for the preparation of polyacrylamide gel were the products of Biorad.

Antitrypsin

cathode buffer 50 mM Tris-glycine, pH 8.9, anode buffer 0.1 M Tris-chloride, pH 8.1 was used. In addition to the above, another preparation of variant Z was obtained by a somewhat different procedure. The latter preparation became accidentally denatured after the completion of the procedure and was used only for analytical purposes (variant Z denatured, column 3, Table 1). Variant M used in this work for comparative purposes was prepared essentially as described above. Analytical Procedures

Purification of Variant Z The selection of the phenotype ZZ plasma donor was based primarily on acid starch gel, crossed immunoelectrophoresis [2,28]. (The phenotyping of plasma was carried out by Dr Diane W. COXof this Institute.) By radial immunodiffusion the plasma contained 0.17 mg rl-antitrypsin/ml and inhibited 0.1 1 mg trypsin/ml. The ACD plasma (250 ml) was dialysed overnight against running cold water to develop a heavy precipitate. After low speed centrifugation at 4 “C a clear supernatant was obtained. All subsequent steps of purification were carried out in the presence of 0.02 sodium azide. In the first step, the plasma was treated in batch on Cellex D in 20 mM acetate buffer, pH 5.8. The dialysed plasma was diluted to twice its original volume in acetate buffer and then mixed with a suspension of about 10 g Cellex D, preequilibrated in the same buffer. After extensive washing of the suspension on a Buchner funnel the protein was eluted with 100 mM NaC1. As a result of this step the purity of the preparation was increased by a factor of about six. Next the partially purified eel-antitrypsin was chromatographed on a QAE-Sephadex column [29] (approximate column dimension 4 x 50 cm) in 50 mM Tris-chloride, pH 8.6 and 100 mM NaCl and eluted in 2.0 1 of buffer by linearly increasing the concentration of NaCl to 400 mM. After equilibration, the inhibitor containing eluate was chromatographed on a Cellex D column [30] in 20 mM acetate buffer, pH 5.8, containing 30 mM NaCl and eluted by linearly increasing the concentration of NaCl to 150 mM in a total of 1.5 1 of buffer. The last two steps of purification consisted of chromatography on DEAE-cellulose [31] (described in Results) and preparative gel electrophoresis [30]. The latter was carried out in a Buchler Polyprep model in a column of 7.5% polyacrylamide gel of 11 cm x 15.8 cm’. The gel mixture contained 0.2% (w/v) bisacrylamide, 0.03 % (v/v) N,N,N‘,N’-tetramethylethylenediamine and 0.08 (w/v) ammonium persulphate in 0.37 M Tris-chloride buffer, pH 8.9. As

Proteins were measured with the Folin reagent [32] using human serum albumin as standard, sialic acid with thiobarbiturate [33], hexoses with orcinol [34], and hexosamines with the Ehrlich reagent [35]. The antigen contents of the different preparations were determined by radial immunodiffusion in Partigen plates, supplied by the Hoechst Co. In all analytical assays, measurements were made at least at four different dilutions of aliquots and concentrations were determined from the resulting slopes. Other Procedures The proteinase inhibition of these preparations was measured by assaying residual protease activity after 15-min incubation with the inhibitor at room temperature. Five different portions of each mixture, ranging from 5 - 25 pl were mixed with 1.5-ml portions of the substrate and after a further 20-min incubation at room temperature, the change in the absorbance of the substrate was measured. Residual trypsin activity was measured with 1.O mM p-tosyl-L-arginine methylester as substrate in 40 mM Tris-HC1, pH 8.05 and 10 mM CaCh and the absorbance was measured at 247 nm. Residual chymotrypsin activity was measured with 2.0 mM N-acetyl-L-tyrosine ethylester in 40 mM Tris-HC1, pH 7.77 and 50 mM CaClz at 240 nm. For the examination of the proteinase - a1-antitrypsin complexes by electrofocusing, the inhibitor was incubated with the proteinase in 20 mM Trischloride buffer, pH 7.2 at room temperature. 5-yl portions of enzyme solutions of varying concentrations were mixed with 40 kg (20 yl) inhibitor and 10-p1 portions of the mixtures were applied to the gel after 10-min incubation. Analytical isoelectrofocusing was carried out in an LKB Multiphore instrument in 5 % (w/v) polyacrylamide gel, containing N,N’-methylenebisacrylamide (0.15 :{ w/v), sucrose (0.125 % w/v) and riboflavin (267 pg/lOO ml gel mixture). The final concentration of the pH 4-6 ampholine in the gel was 5 % (v/v). For the electrofocusing range of pH 5.0 - 8.5, 2.5 ”/, (v/v) of each of the ampholines of pH 5 - 7 and pH 7-9 were used. As anode electrolyte 1.0 M

605

A . Hercz and M . Barton

H3P04and cathode electrolyte 1.O M NaOH was used. The duration of migration was approximately 3 h. Desialysation of glycoproteins with protease-free neuraminidase was carried out essentially by the method of Mohr et al. [36]. 200-pg portions of either variant M or Z were treated at room temperature with varying amounts (12.0 U - 40.0 U) of neuraminidase in 50 mM sodium acetate buffer, pH 5.5 containing 150 mM NaCl and 9 mM CaC12. The final volume of the incubation mixture was 200 p1 and 25 pl samples were drawn at 4-min intervals for the determination of free sialic acid. After 30-min incubation 10-pl samples were analysed by isoelectrofocusing.

0.3

-

4

0.2

?

RESULTS 0.1

After purification on the Cellex D column (see Methods), the variant-Z-containing fraction was chromatographed on a DEAE-cellulose column as illustrated in Fig. l. Measurement of the trypsin-inhibiting capacities of the fractions revealed the presence of three distinct peaks, the first two of which overlapped with immunologically active a1-antitrypsin. (In Fig. 1 the three fractions with the highest trypsin inhibition are indicated by arrows. The rest of the results is not plotted.) The fractions were pooled as indicated (Fig. 1) and then pool A was further purified by preparative electrophoresis in polyacrylamide gel (see Methods). Only the first and second peak, obtained by this procedure, were collected. The first peak eluted from the preparative polyacrylamide gel column, containing 6.7 mg variant Z was homogeneous (sample 2, Fig. 2) and its proteaseinhibiting capacity and antigen content was comparable to that of variant M (lines 22 - 24, Table 1). The two most troublesome contaminants in al-antitrypsin preparations, albumin and al-acid glycoprotein, were not detected in this protein by immunoelectrophoresis using monospecific antisera. The second peak emerging from the preparative gel column was also homogeneous (sample 3, Fig.2) and moderately inhibited trypsin, but did not react with antiserum to al-antitrypsin. We have not so far identified and characterized the protein in the second peak. Pool B in Fig.1 was also purified by preparative electrophoresis in the same way as pool A to yield 1.5 mg pure al-antitrypsin. (The latter preparation was not used in the subsequent experiments and the term variant Z hereafter will designate the protein derived from pool A.) The starting material, 250 ml plasma, contained 42.3 mg crl-antitrypsin (0.17 mg/ml) and 14.7 g protein (58.8 mgiml). From this, the 8.2 mg purified al-antitrypsin (6.7 mg from pool A and 1.5 mg from pool B) corresponds to a 350-fold purification, at a yield of 19 %.

f.

0 0

150

200 Fraction number

250

Fig. 1. C'liromatogruphy oj the Z variunr on u DEAL~-celIulo.sc~ column in 5 m M sodium phosphate buffer, p H 6.4, containing 50 m M NaCI. The preparation used in this step was obtained by the preliminary purification described in Methods. Approximate column dimension: 2 cm x 70 cm. Elution was carried out with a linear gradient of NaCl (1.6 1) ranging in concentration from 50 m M - 150 mM. The arrows indicate the positions and relative heights of the three peaks found by trypsin inhibition assay. A and Bin brackets indicate the pooling offractions. Absorbance(-----); [NaCI] (-) antigen content (0-0);

The purified variant Z was compared with an authentic preparation of variant M and with the sera of the respective donors by electrofocusing as illustrated in Fig.3. The isoinhibitors in the purified proteins coincided with their counterparts in the corresponding serum aliquots (compare sample 1 with 2 and 3 with 4). In agreement with earlier observations (16- 19) made on whole serum, the components of variant Z (sample 2) as a group were cathodal to, but partially overlapped with those of variant M (sample 3). Next we measured the time-course of desialysation of variants M and Z with varying amounts of neuraminidase (0.06- 0.20 U/pg al-antitrypsin, see Methods). The results were consistent with a first-order reaction and the liberation of sialic acid rapidly declined after the initial 15-20 min of incubation. After 30-min incubation, the concentration of free sialic acid, up to a limit, was roughly proportional with the amount of neuraminidase added and 0.088 U/pg a1 -antitrypsin released 94 % of the glycoprotein-bound sialic acid. Further increasing the neuraminidase up to 0.2 U/pg sample did not result in measurable change. (Proteolytic activity could not be detected in the neuraminidase preparation used in this experiments.)

606

Antitrypsin

1

2

3

4

Fig. 2. Analytical electrophoresis in 7.5 ‘i; poljucrylamide gel of samples obtained by preparative electrophoresis in polyucrylarnide gel. Gel 1 : sample before preparative electrophoresis; Gel 2: the 1st peak (variant 2 ) and Gel 3: the 2nd peak, eluted from the gel column. Migration from top to bottom and anode is at the bottom

Fig. 3 . lsoelectryfocusing oJ purified r l-unlitrjpsnl preparotions and serum aliquots in polyacrylumide gel between p H 4.0-6.0. Sample 1 : serum Pi phenotype ZZ; sample 2: purified variant Z; sample 3 : purified variant M ; and sample 4 : serum Pi phenotype M M . Migration from bottom to top and anode is at the top

On the basis of these measurements, aliquots of variants were treated for 30 min with 0.06 U/pg c11antitrypsin (corresponding to 64 % desialysation) and 0.15 U/pg al-antitrypsin (1.7 times the amount causing 94 % desialysation) and then electrofocused as illustrated in Fig. 4. After partial desialysation (samples 5 and 6) the samples shifted from their original to cathodal positions, as expected and increased level of desialysation resulted in a less pronounced but detectable additional change (samples 3 and 4). A more important aspect of this experiment was that unlike in the controls, the isoinhibitors in the different variants after desialysation completely coincided (compare sample 3 with 4). However, the balance of the isoinhibitors in the desialysed variants was different, as variant Z (sample 4) contained a larger proportion of cathodal and smaller proportion of anodal components than variant M (sample 3). To further elucidate the relationship between the isoinhibitors of these variants, we examined their complexes with trypsin (Fig. 5 ) and chymotrypsin (Fig. 6) by electrofocusing between pH 5.5- 8.0. At least five individual isoinhibitors in both variant M

(sample 3 and 5 , Fig. 5 ) and variant Z (sample 4 and 6, Fig. 5) formed completely overlapping complexes with trypsin (compare sample 3 with 4 and sample 5 with 6). In line with the previous finding (Fig.4) albeit less obviously, in variant Z (sample 6, Fig.5), the proportion of the most cathodal trypsin-inhibitor complex was increased and the most anodal decreased relative to variant M (sample 5, Fig. 5). A complete overlap between the complexes of these variants with chymotrypsin is illustrated in Fig. 6. The amino acid and carbohydrate compositions of variant M and variant Z are listed in Table 1. Both variants contain approximately 400 amino acid residues per 50000 g protein and an appreciably high proportion of aspartic (asparagine) and glutamic (glutamine) acid residues, characteristic of glycoproteins in general. The amino acid composition of variant M on the whole agrees reasonably well with earlier data [26,29,31,37,38]. The quantitatively most important difference between these compositions was a 36% increase in the glycine content of variant Z (line 6, Table 1). According to the data in columns 2

A. Hercz and M. Barton

607

.

1

2

4

3

5

6

7

8

Fig. 4. Isoelectrofocusing ( p H 4.0 -6.01 qfpurqyed ccl-untitrjpsin prqxuutioris after various treutnzents ~ t i t hneuraniiniduse. Purified variant M : samples 1, 3, 5 and 7. Purified variant Z : samples 2, 4, 6 and 8. Before treatment with neuraminidase: samples 1, 2, 7 and 8. After treatment with 0.06 U of neuraminidase/pg glycoprotein: samples 5 and 6 . After treatment with 0.15 U of neuraminidase/pg glycoprotein: samples 3 and 4. Anode is at the top

1

5

6

Fig. 5. lsocleclrqfocusiiig ( p H 5.5-8.0) oj ccl-antitrypsin after treatment with trypsin. Samples 1, 3 and 5 : variant M ; samples 2, 4 and 6 : variant Z. Portions of 40 pg inhibitor were treated with the following amounts of trypsin: Samples 1 and 2: 0 pg (controls); samples 3 and 4: 16.0 pg; samples 5 and 6: 8.0 pg. (The conditions of incubation are described in the Methods.) Anode is at the top

Fig.6. I,~oelectrofocu.sin~( p H 5.5-8.0) of r,-antitrl,psin after treatment with cc-chyrnotrypsin. Sample 1 : variant M ; samples 2 and 3: variant Z. Portions of 40 pg inhibitor were treated with the following amounts of al-chymotrypsin: Samples 1 and 2: 5.0 pg, sample 3: 0 pg, (For additional details see Methods.) Anode is at the top

and 3, Table 1, small differences in respect of several residues existed even between the two variant Z preparations. Because of this, for the time being we do not attach significance to additional minor differences between the figures in columns 1 and 2, Table 1. No appreciable difference was found between the carbohydrates of these variants (lines 19- 21, Table 1).

DISCUSSION The first step in the purification of these variants, batch purification on Cellex D was used for some time in this laboratory with consistent success. We introduced this method not as much for the actual overall purification as for the facility to extensively wash the

608

Antitrypsin

Table 1. Amino acid and carbohydrate composition, antigen content 2,-antitrypin Amino acids and carbohydrates are expressed as number of residues/mole protein. Amino acid residues were calculated from 24-h, 48-h and 72-h acid hydrolysates. The hydrolysates were analyzed on a Technicon Autoanalyzer (9TSM). The inhibitions are expressed as mg proteinase inhibited/mg inhibitor. The data are not adjusted to the number of active sites in the proteinase preparation. Absorbance is expressed as A280 in 1 cm with 0.1 protein and prorrase inhibiting capacity of

Variant M

Variant Z

Variant Z (denatured)

residues/mol protein ~~~~~~

1. Aspartic acid 2. Threonine 3. Serine 4. Glutamic acid 5. Proline 6. Glycine 7. Alanine 8. Valine 9. Methionine 10. Isoleucine 11. Leucine 12. Tyrosine 13. Phenylalanine 14. Lysine 15. Histidine 16. Arginine 17. Tryptophan” I 8. Cysteine 19. Hexose 20. Hexoseamine 21. Sialic acid _ _

44.82 28.50 22.47 50.20 17.27 23.74 26.61 26.75 7.46 20.60 45.03 6.89 28.17 35.05 12.52 7.46 1.80 1.74 19.70 13.55 5.89

42.78 28.11 21.90 49.98 18.80 32.24 25.77 26.50 7.15 20.22 44.35 6.16 28.07 33.76 13.41 7.51 1.88 1.72 22.43 13.58 6.39

44.77 28.17 22.49 53.00 16.05 34.52 28.51 26.24 7.15 20.12 43.15 5.89 25.83 33.73 12.52 7.75

97

91

-

1.79 19.06 12.24 6.04

~~~~

‘?(,

22. Antigen content _ _ ._

_

protein

98

~~-

~

~

~~~

~

mg/mg inhibitor

23 Trypsin inhibition 24 Chymotrypsin inhibition _ _ ~ _

0.47

0.53

< 0.10

-

25. Absorbance a

Determined from ultraviolet spectrum. Measured as cysteic acid.

cellulose-bound protein before elution. From several observations we believe that by its application we remove, partly at least, certain plasma components (lipoproteins?) that otherwise would interfere with the later stages of the purification. The rest of the purification steps consisted of established procedures. After elution from the DEAE-cellulose column (Fig. I), variant Z was found in association with two distinct protein peaks. We have made comparable observations on several occasions in the purification

of variant M as well. As we have not so far characterized the Z variant from pool B, its relation to the protein from pool A remains to be elucidated. Several years ago Schultze et al. [39] suggested that al-antitrypsin may have a tendency to dimerise during purification. Recently Myerowitz et al. separated two physically distinct al-antitrypsin fractions from human serum [40] and antigenically distinguishable fractions from mouse serum [41] by preparative electrophoresis in polyacrylamide gel. The observation in Fig. 1 may be related to the latter findings [39-411, but it simply could be a consequence of complex formation between ccl-antitrypsin and various other serum proteins. The final product of the purification was homogeneous by several independent criteria, had a high protease inhibiting potency, antigenic activity and an isoinhibitor composition similar to the one in the plasma of the donor. On these grounds the preparation was judged suitable for these studies. In several experiments (Fig.4, 5, 6) variant Z was compared with variant M by electrofocusing after various treatments, to find a possible difference between their contents of charged residues. For the examination of the asialoproteins, more than 94 ‘%; of the glycoprotein-bound sialic acid could not be removed even when the amount of neuraminidase was increased from 0.088 U/pg sample to 0.200 Ujpg sample. Considering the error of the measurement, for all practical purpose this may correspond to complete desialysation. While variants M and Z in their native state occupied only partially overlapping positions in the electrofocusing field, after desialysation the overlap became complete (Fig. 4). This could be interpreted to mean that native variant Z contained fewer sialyl residues than native variant M. Since this is not supported by our chemical data (line 21, Table l), we submit an alternative explanation. We assume that similarly to other glycoproteins, s(1 -antitrypsin also has a strong tendency to form aggregates, which in turn would interfere with the correct electrofocusing expression of this protein. The commonplace observation that the duration and conditions of storage drastically affect the electrophoretic properties of al-antitrypsin supports this argument. On these premises, less than complete overlap of the isoinhibitors would be seen before desialysation, because the increased proportion of the cathodal components in variant Z would lead to the formation of aggregates cathodal to those in the M variant. Concomittantly with the liberation of sialyl residues, treatment with neuraminidase would result in dissociation of the aggregates thereby establishing complete overlap. The observations that even after partial desialysation the resolution of these proteins greatly improved and near complete overlap became established, support this argument.

609

A. Hercz and M . Barton

In a recent report by Talamo et al. [42], differences between the isoelectric points of the phenotypes FF, MM, and ZZ of al-antitrypsin remained after desialysation. Cox [43] reporeted that the asialo Z variant was only slightly cathodal to asialo M protein on agarose electrophoresis at pH 8.6. As the previous authors [42- 431 carried out these experiments with whole serum, their results may not be comparable with the present results. The large amount of glycoproteins, other than rl-antitrypsin in the serum may have an adverse effect on the process of desialysation and the ensuing observation by electrofocusing and electrophoresis. After the completion of this investigation, two publications appeared indicating that in the Z variant, at least one [44] and possibly two [45] glutamyl residues were replaced by positively charged amino acids (lysine and glutamine). From those results [44,45] the net positive charge on the asialo-Z variant relative to the asialo-M species would be increased by two or possibly four units. Judging from the remarkably large cathodal shift after desialysation (Fig. 4), the sensitivity of this electrofocusing system would be sufficiently high to detect a difference of at least four units (and possibly as little as two units) between the net charges of the variants. Thus the complete overlap of the asialo-variants (Fig. 4), seems to conflict with the earlier observations [44,45]. One possible explanation of this conflict may be that the amino acid substitutions above are balanced by a hitherto unrecognized opposing change(s) of charged residues. The observations in F i g 3 and 6 prove that both of the M and Z variants contain at least five individual components, capable of forming distinct complexes with trypsin and chymotrypsin. The fact that the corresponding complexes of these proteins completely overlapped with each other seems to corroborate the previous evidence (Fig. 4), albeit not without qualification. It is conceivable that the net charges of the two proteins are different, but the segments of the molecules in which the corresponding difference is localised become eliminated from the complex through a proteolytic cleavage(s) in the interaction of the variants with the proteinase. There are indeed observations [37,46] indicating that x1-antitrypsin is cleaved by trypsin in the course of the inhibition reaction, but this is less likely to take place with chymotrypsin [46]. However, an unconditional interpretation of these observations (Fig.5 and 6) is subject to a better understanding of the interaction of al-antitrypsin with these proteases. The presence of al-antitrypsin exclusively in the rough endoplasmic system of the hepatocytes in the deficiency state, strongly suggests that in this condition, primarily the protein moiety of the glycoprotein is affected (see Introduction). The increased glycine content of the Z variant (line 6, Table 1) is in

line with this view. A similar increase of 8 glycine residues was reported earlier by Chan et al. [47], but unlike us, they also found an increase of four arginyl residues in the composition of the Z protein. The glycine content of the Z protein isolated from liver inclusions was increased by four or five residues in addition to appreciable differences in the arginine and phenylalanine content [26]. In two other reports no difference was found between the amino acid compositions of the M and Z protein [45,48]. Unlike other authors [45,47,48], we could not find an appreciable difference between the carbohydrates of these variant (lines 19-21, Table 1) in spite of repeated and careful assay. Among the authors, who did find a difference, there is a substantial disagreement about its extent. According to Jeppsson et al. (quoted in [44]) and Yoshida et al. [45], only the sialyl residues were appreciably decreased (by 1 or 2 residues) in the Z variant but the rest of the carbohydrates was not affected. In contrast, Chan et al. [47] and Miller et al. [48] reported that in the Z protein all carbohydrates were decreased by approximately 25 %. From the survey of the results in the foregoing, there is a conflict between the observations of different authors. Some of these discrepancies may be due to experimental errors, but individual variations, particularly in the carbohydrate composition of the Z variant cannot be ruled out at present. Depending on its duration and gravity, the disturbance of the biological functions in the liver of antitrypsin-deficient persons may affect differently the pathway leading to the completion (glycosylation) and transport of this protein. Another possibility to consider is that some genetic differences may exist between persons of ZZ genotype, but they cannot be detected by the methods at present available. The authors are grateful to Prof. David M . Goldberg for helpful suggestions in the writing of the manuscript. This work was supported by a grant to Albert Hercz from the Medical Research Council of Canada.

REFERENCES 1. Kueppers, F. & Black, L. F. (1974) A m . Rev. Rc,spir. Disease,. 110, 176-194. 2. Fagerhol, M . K. & Laurell, C. B. (1967) Clin. C'hinl. Acta. 16. 199 - 203. 3. Fagerhol, M . K . & Gedde-Dehl, T. (1969) Hum. Hercd. IY. 354- 359. 4. Fagerhol, M. K. (1967) Acra Pathol. Microbial. Scuncl. 70, 421 -428. 5. Talamo, R. C., Langley, C. E., Reed, C. E. & Makino. S. (1973) Science (Wash. D.C.) 181, 70-71. 6. Altay, C., Fagerhol, M. K., Erdogan, N. & Say, B. (1973) N . Engl. J . Med. 289, 754. 7. Martin, J. P., Vandeville, D. & Ropartz, C. (1973) Lancet, 2, 845. 8. Laurell, C. B., Sveger, T. & Ljunggren, C. G. (1974) Acra Prdiatr. Scand. 63, 8 5 5 - 857.

610 9. Martin, J. P. & Ropartz, C. (1975) Rev. Fr. Tran.Tfus. ImmunoHematol. 18, 27 - 45. 10. Cox, D. W. (1975) Protides B i d . Fluids Proc. Colloq. Bruges, 23,375 - 378. 11. Lieberman, J., Guidulis, L. & Klotz, S. D. (1976) Am. Rev. Respir. Dis. 113, 31-36. 12. Sharp, H. L. (1971) Hosp. Pract. 6, 83-96. 13. Lieberman, J., Mittman, C. & Gordon, H. W. (1972) Science ( Wadi. D.C.) 175, 63 - 65. 14. Palmer, P. E., DeLellis, R. A. 81 Wolfe, H. J. (1974) A m . J . Clin. Pathol. 62, 732-739. 15. Gordon, H. W., Dixon, J., Rogers, J. C., Mittman, C. & Lieberman, J. (1972) Human Pathol. 3, 361 - 370. 16. Laurell, C.-B. (1965) Scand. J . Clin.Lab. Invest. 17, 271 -274. 17. Kueppers, F. &Beam, A. G. (1966) Science (Wash. D. C . ) 154, 407-408. 18. Allen, R. C., Russel, A. H. & Talamo, R . C. (1974) Am. J . Clin. Pathol. 62, 732- 739. 19. Lebas, J., Hagem, A . & Martin, J. P. (1974) C. R. Hebd. Seances Acad Sci. Paris, 278, 2359-2360. 20. Bell, 0. F. & Carrel], R. W. (1973) Nature (Lond.) 243, 410411. 21. Cox, D. W. (1973) Lmzcet, 844, 2. 22. Weiser, M. M., Lamont, J . T. 81 Walker, W. A. (1975) N . Engl. J . Med. 292, 205 - 206. 23. Pricer, W. E. & Ashwell, G. (1971) J . B i d . Chem. 246, 48254833. 24. Morell, A. G., Gregoriadis, G. & Scheinberg, I. H. (1971) J . B i d . Chem. 246, 1461- 1467. 25. Kuhlenschmidt, M. S., Yunis, E. J., Iammarino, R . M.. Turco, S. J., Peters, S. P. & Clew, R. H . (1974) Lab. Invest. 31, 41 3 -419. 26. Jeppsson, J.-O., Larsson, C. & Erikksson, S. (1975) N . Enx. J . Med. 293, 576- 579.

Antitrypsin 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

Cox, D. W . (1975) Lab. Invcst. 32, 777. Fagerhol, M . K. (1968) Ser. Hematol. I , 153-161. Crawford, 1. P. (1973) Arch. Biochem. Biophys. 156, 215-222. Murthy, R. J. & Hercz, A. (1973) FEBS Lcitt. 32, 243-246. Pannel, R., Johnson, D. & Travis, J. (1974) Biocheniisti:i,, 13, 5439 - 5445. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J . Biol. Chem. 193, 265-275. Warren, L. J. (1959) J . B i d . Cheni. 234, 1971 - 1975. Weimer, M. E. & Meshin, J . R. (1953) Anz. Rev. Tubercir/o,sis. 68,594 - 602. Rimington, C. (1940) Biocl7c~m.J. 34, 931 -940. Mohr, E. & Schramm, G. (1960) Z. Natuyforsch. 153,568- 574. Moroi, M . & Yamasaki, M. (1974) Biochim. Biophys. Acta, 359. 130-141. Kress, L. F. & Laskowski, M. Sr. (1973) Prep. Biochem. 3, 541 - 552. Schultze, H. E.,Heide, K. & Haupt, A. (1962) Klin. Wochenschr. 40,427 - 429. Myerowitz, R. L., Handzel, 2 . T. & Robbins, J . B. (1972) Clin. Chim. Acta, 39, 307-317. Myerowitz, R. L., Chrambach, A , , Rodbard, D. & Robbins, J. B. (1972) Anal. Biocliem. 45, 394-409. Talamo, R. C., Alpert, E. & Langley, C. E. (1975) Pcdiatr. Res. 9, 123- 126. Cox, D. W. (1975) Am. J . Hum. Genct. 27, 165-177. Jeppsson, J.-0. (1976) FEBS L e t t . 65, 195- 197. Yoshida, A,, Lieberman, J . , Gaidulis, L. & Ewing, C . (1976) Proc. Nut Acad. Sci. U . S . A . 73, 1324-1328. Hercz, A. (1974) Eur. J . Bioclzeni. 49, 287-292. Chan, S. K. & Rees, D. C . (1975) Nature (Lorzd.) 255, 240241. Miller, R . R., Kuhlenschmidt, M. S., Coffee, C. J., Kuo, I. & Clew, R. H. (1976) J . Biol. Chem. 251, 4751 -4757.

A. Hercz, Research Institute, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada, M 5 G 1 x 8 M. Barton, Biochemistry Research Department, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada, M 5 G 1 x 8