- Email: [email protected]
Carrier detection in cystic fibrosis
Carrier detection in cystic fibrosis We evaluated four methods purported to distinguish between individuals homozygous or hetep?ozygous for cystic fibrosis (CF) and normal controls." (1) detection of a protein in the serum by isoelectric focusing at pH 8.5, i2) detection of a lectinlike factor in the serum by hemagglutination, (3) isolation of CF-lectin from the serum by .affinity ehromatog~'aphy, and (4) measurement of MUGB-reactive proteases in the plasma. The results were disappointing. The detection of CF protein by isoelectric focusing was unreliable; it could be identified in only 46 % of heterozygotes and 66 % of homozygotes, with a false positive rate of 17 %. Detection of a lectinlike factor by hemagglutination was also found to be unreliable and irreproducible. The lectin isolated by affinity chromatography was not specific for the CF gene. No significant differences were found in the MUGB titers of the three populations tested. However, low titers (MU <200 nmol/ml) were found in 33 % of homozygotes and heterozygotes and in 17% of normal controls. We conclude that none of these methods is suitable for carrier detection in cystie fibrosis. (J PEDIATR 106:913, 1985)
A. Rauf Qureshi, Ph.D., and Hope H. Punnett, Ph.D. Philadelphia, Pennsylvania
No TEST is presently available for detection of cystic fibrosis carriers. Although many claims for methods of heterozygote detection have been advanced, few have been confirmed. We have evaluated four reputed markers of the CF gene for their reliability: (1) the cystic fibrosis protein detectable in serum by isoelectric focusing at pH 8.5!; (2) the presence of a lectinlike factor in the serum of CF homozygotes and heterozygotes that agglutinates mouse red cells2; (3) the CF-lectin isolated by affinity chromatography on fructose-conjugated sepharose 6B column3; and (4) the presence of reduced levels of MUGB-reactive proteases in the plasma of CF genotypes. 4,5 We evaluate d these methods for reproducibility in hopes that one alone or a combination of two or more might prove suitable for carrier detection. From the Department of Laboratories, Genetics Section, St. Christopher's Hospital for Children, and the Department of Pediatrics, Temple University School of Medicine. Supported in part by Grant A M 26606 from the National Institutes of Health and by the Cystic Fibrosis Foundation. Submitted for publication Aug. 27, 1984; accepted Dec. 4, i984. Reprint requests." A. Rauf Qureshi, Ph.D., Department of Laboratories, Genetics Section, St. Christopher's Hospital for Children, 5th and Lehigh Ave., Philadelphia, PA 19133.
METHODS Venous blood from patients with CF, obligate heterozygotes, and normal healthy adult controls was collected with informed consent. The controls and heter0zygotes were age matched (range 22 to 54 years). The age range of the patients was 1 to 29 years. The clinical diagnos!s of CF was confirmed by sweat electrolyte analyses. All patients were being seen at the CF clinic at St. Christopher's Hospital
See related article, p. 930. CF MU MUGB
Cystic fibrosis 4-Methylumbelliferone 4-Met hylumbelliferylguanidinobenz0ate
for Children. At the time when blood specimens were taken, most of the patients were receiving supplementarY pancreatic enzymes and antibiotics orally. E!eetrofocusing of CF protein. The concentration of IgG in each sampl e was determined by single radial immunodiffusion using commercial kits (Kailestad Laboratories, chaska, Minn.). Isoelectric focusing in thin-layer polyacrylamide gel using the LKB Muttiphor (Model 2117; LKB Produkter,
The Journal of P E D I A T R I C S
913
9 14
Qureshi and Punnett
The Journal of Pediatrics June 1985
1
2=
3
4
5
6
7"
8 ~
9
Fig. 1. Isoelectric focusing pattern of serum proteins obtained in pH 2.5 to 10.0 gradient. Serum volumes containing 300 or 600 ug lgG were applied on 3 mm pads at anode. Samples were prefocused for 30 minutes at 350 V. then focused for 90 minutes at 35 W. Samples 2.4. 6. and 8 from CF homozygotes, samples 1.3.5.7. and 9 from normal controls. Samples 7. 8, and 9 are same as 3, 2, and 1, respectively, except that 600 ~g IgG was used for samples 7 through 9. Bars indicate presence of CF protein bands. Only samples containing 300 ~zg IgG were considered for presence or absence of CF protein bands. Table I. Presence of CF protein in serum, as determined by isoelectric focusing
In Controls CF heterozygotes CF homozygotes
29 33 64
CF protein band(s) Present Doubtful Absent n
%
n
%
n
5 15 42
17 45.5 65.5
4 10 14
14 " 20 30 8 22 8
% 69 24 12.5
Bromma, Sweden) was performed according to the method of Wilson et al. 6 with unpublished modifications (personal communication, 1980). DeteCtion of CF lectin. Lectinlike activity in the serum w a s m e a s u r e d first according to the earlier methods described by Lieberman and co-workers:. 7 and also according to their modified method (personal communication, 1981). Isolation of CF leetin. Affinity chromatography was performed on fructose-Sepharose 6B column according to the described method of Nordstrom et al? MUGB-reaetive proteases. The titer of MUGB-reactive proteases was measured in citrated plasma of C F homozygotes, heterozygote s, and normal controls by previously described methods.,4 5 Nonspecific hydrolysis of M U G B was corrected by carrying out the reaction in the presence and absence of benzamidine, a competitive inhibitor of serine proteases. M U G B reactivity was measured in unactivated plasma 0nly. ~
!soelectric focusing of activated and unactivated plasma was carried out in 5% polyacrylamide gel, with 0.75% Ampholine, pH 5 to 7, and then stained for M U G B reactivity? RESULTS CF protein. Samples were considered positive for C F protein if a single or a doublet band or a sharp frontal edge was detectable in the most cathodal centimeter of the gel (pI 8.5 to 8.7) (Fig. 1). The C F protein was Present in 46% of heterozygotes and 66% of homozygotes for the C F gene and in 17% of normal controls (Table I). CF leetln. The test was performed in 29 patients with CF, five obligate heterozygotes, and 13 normal controls. Initial results were encouraging. Lectinlike activity was found i n t h e sera from 27% of C F homozygotes and 45% o f heterozygotes, but was present in only 4% of control sera. However, contradictory results were obtained when there was a Change in laboratory personnel, and our earlier results could not be reproduced. Isolation of CF leetin. We were able to desorb a protein peak from the serum of normal controls and C F heterozygotes. A similar peak could also be desorbed from C F homozygotes. The test was performed in four controls, three patients with CF, and four obligate heterozygotes. Lectin was present in all of the sera. Therefore, the lectin peak was not found to be specific for the C F gene (Fig. 2). MUGB-reaetive proteases. Initially, we measured M U G B reactivity on activated and unactivated plasma
Volume 106 Number 6
Carrier detection in cystic fibrosis
2.0
9 15
'
1-6r
o
Ii
N
I I
--
1-2
~
0"8
I, 't
.
iiI
0
~ 0
If'
14
t
1 L
28
42
56 Volume (ml)
112
126
140
^
-
^
154
Fig. 2. Affinity chromatography on fructose-conjugated Sepharose 6B column (0.9 • 15 cm). Column was equilibrated with initial buffer (0.9% saline solution containing 1 mM MgC12, pH 6.0). Four milliliters of serum was diluted with initial buffer and applied to column. Fructose-specific lectin Was eluted (arrow) with initial buffer containing 10 mM EDTA. Effluents were monitored by measuring optical density at 240 nm, using EDTA buffer-as blank..e, Heterozygous serum; ~ normal control. and confirmed (data not shown) the observations of Rao et al. 4 that there was no significant difference in the titers of activated and unactivated plasma. Subsequently, M U G B reactivity was measured in unactivated plasma only. Our results (Table II) show that there was no significant difference in titers or specific activity in plasma from normal controls, heter0zygotes, and homozygotes We did not find any correlation between M U G B reactivity, albumin levels ( X + SD 41.6 _+ 6.4 mg/ml), and a g e of patients with CF. Our data also showed that low M U G B titers ( M U < 200 n m o l / m l plasma) were present in 17% of normal controls, 33% of heterozygotes, and 31% of homozygotes. There was an extensive overlap among the three populations. We were also unable to confirm the differences in the number of bands obtained after isofocusing and staining with M U G B in the plasma of C F genotypes and controls. DISCUSSION The primary goal of this study was the evaluation of published tests purporting to distinguish unaffected individuals who are heterozygous for the C F gene from those who are homozygous normal or who are affected (homozygous for the C F gene) and to see if one method alone or a combination of two or more might prove suitable for carrier detection. Inasmuch as only one method showed some promise, the second part of the study was not done. C F lectin is reported to occur in both C F homozygotes and heterozygotes. The agglutinating activity in C F serum is inhibited by fructose, fucose, and antibiotics. The
Table II. Plasma M U G B activity
(nmol MU/ml)
Specific activity (nmol MU/mg total protein)
232 _+ 38 174 to 285
3.7 _+ 0.5 3.1 to 4.5
225 _+ 44 146 to 293
3.6 _+ 0.7 2.5 to 4.8
214 _+ 46 115 to 304
3.3 -+ 0.8 1.9to5.0
Titer
n Normal controls Mean _+ SD Range Heterozygotes Mean _+ SD Range Cystic fibrosis Mean_+ SD Range
12
24
23
non-CF activity found in some controls is inhibited by mannose. 2 The test as described is simple. However, it is very subjective because there is no sharp end point in scoring agglutinatio n . We have found it difficult to obtain consistent results in successive tests using the same specimen. We did a blind study using specimens collected and coded by the Pulmonary Department, St. Christopher's Hospital for Children. A change of personnel in the laboratory led to contradictory results, reinforcing our concerns about subjectivity. This test was found to be unreliable and unsuitable for screening purposes. Using affinitY column chromatography on fructoseSepharose 6B, we have found the fructose-specific lectin to be present in all" sera tested, including that of normal Controls and C F l~eterozygotes and homozygotes. This finding is contrary to the observations o(]'qordstrom et al., 3 who found that the leetin peak was absent from the sera of controls. However, the authors also pointed out that the
9 16
Qureshi and Punnett
lectin peak cross-reacted with the anti-human serum, suggesting that the lectin was not unique for the CF gene. Lieberman and Kaneshiro9 found a low molecular weight (1000 daltons) lectin in the serum of controls and CF heterozygotes and homozygotes. Only lectin from patients with CF and carriers showed any hemagglutination activity. Similar findings have also been made on CF factors ~~ that are ubiquitous in serum and urine but possess biologic activity only when isolated from CF genotypes. Reduced levels of MUGB-reactive proteases in CF homozygotes and heterozygotes were reported by WalshPlatt et al. ~ We did not find significant differences in the titers of MUGB-reactive proteases in the plasma of CF homozygotes, CF heterozygotes, and controls. However, we did find low titers (MU <200 nmol/ml) in approxi 7 mate!y 33% of CF homozygotes and heterozygotes and in 17% of normal controls. The extensive overlap among the three groups renders this test unsuitable for carrier detection. MUGB has been shown 13 not to meet the requirements of an active site titrant under the conditions described for this assay. Although we did not find any correlation between the levels of albumin and the MUGB reactivity in the plasma of patients with CF, others 14 have found plasma albumin, rather than plasma proteases, to be responsible for MUGB reactivity, because purified albumin was shown to possess MUGB hydrolysing activity. However, the possibility of MUGB-reactive material (e.g., a polypeptide produced as a result of catabolism of arginine esterase in the circulation) being tightly bound to albumin cannot be ruled out. Support for this concept was provided by the isofocusing pattern of MUGB-reactive material in the plasma, inasmuch as identical patterns were obtained in all three populations. This is consistent with the fact that albumin also focuses in the same pI region (5.2 to 5.5) 15 as MUGB-stainable proteases (pI 5.1 to 5.4). 8 Therefore, the determination ~of MUGB reactivity or isofocusing of MUGB-stainable bands for the purpose of carrier detection is not recommended unless--and until--the putative enzyme is identified and characterized: Isoelectric focusing in polyacrylamide gel of serum proteins has been claimed to be a useful method ~for carrier detection. We have confirmed the presence of CF protein in the sera of some individuals heterozygous or homozygous for the CF gene. We have found the CF protein in approximately 46% of heterozygotes and 66% of homozygotes for the CF gene. This is a lower figure than that reported by WilsonI and by others. ~6.t7 Findings similar to ours have also been reported. 18,~9 The frequency of CF protein in our control subjects was 17%, which is higher than the expected carrier frequency in white populations. Although the isofocusing technique is straightforward,
The Journal o f Pediatrics June 1985
requiring no unusual-steps, the bands are often equivocal. They are not always detectable, perhaps because of low levels and lack of sensitivity of the isofocusing techniques. Morever, the method requires some degree of subjectivity in evaluation of results, and is therefore unreliable and unsuitable for population screening, although it may have a role in family studies. The CFP region of polyacrylamide gel has been used to raise antibody in mice2~and in guinea pigs. 2~ Both antisera were reported to distinguish quantitatively between CF homozygotes, heterozygotes, and controls. Bullock et al? 2 failed to produce specific antibody in any species other than guinea pig despite considerable efforts. We have attempted to raise antibody in guinea pigs according to the methods of Bullock, but have not been successful. We conclude that all four methods tested are unsuitable for detection of CF carriers. At present there is no known definite marker for the CF gene that can be used for screening: We thank Lucinda DeMarco and Gregory Higbee for technical assisance; and Dr. Judy Palmer, Dr. Daniel Schidlow, and the staff of the Cystic Fibrosis Center for collecting blood samples. Our special thanks to the patients with CF and 9 parents who donated blood for this study. REFERENCES
1. Wilson GB, Fudenberg HH: Studies on cystic fibrosis using isoelectric focusing. I. An assay for detection of cystic fibrosis homozygotes and heterozygote carriers from serum. Pediatr Res 9:635, !975. 2. Lieberman J, Costea NV, Jakulis VJ, Kaneshiro W: Detection of a lectin in the blood of cystic fibrosis homozygotes and heterozygotes. Trans Assoc Am Physicians 92:121, 1979. 3. Nordstrom OV, Blomfield J, Brown JM: Isolation and characteristics of a fructose specific lectin from cystic fibrosis homozygote and heterozygOte serum. In Sturgess JM, editor: Perspectives in cystic fibrosis. Proceedings of the Eighth International Congress on Cystic Fibrosis, Toronto, Ontario Canada, 1980, p 169. 4. Rao GJS, Walsh-Platt M, Nadler HL: Reaction of 4methylumbelliferylguanidinobenzoatewith proteases in plasma of patients with cystic fibrosis. Enzyme 23:314, 1978. 5. Walsh-Platt M, Rao GJS, Nadler HL: Protease deficiencyin plasma of patients with cystic fibrosis. Enzyme 24:224, 1979. 6. WilsonGB, Arnaud P, Fudenberg HH: Improved method for detection of cystic fibrosis protein in serum using the LKB Multiphor electrofocusing apparatus. Pediatr Res 11:986, 1977. 7. Lieberman J, Kaneshire W, Costea N: Characteristics of a new screening test for detecting the cystic fibrosisgene: Assay of a serum lectin. In Sturgess JM, editor: Perspectives in cystic fibrosis. Proceedings of the Eighth International Congress on Cystic Fibrosis, Toronto, Ontario, Canada, 1980, p 308. 8. Rao GJS, Nadler HL: Arginine esterase in cystic fibrosis of the pancreas. Pediatr Res 8:684, 1974.
Volume 106 Number 6
9. Lieberman J, Kaneshiro W: Detection of a low molecular weight CF-lectin cofactor in serum by G-25 gel filtration. CF Club Abst, 1981, p 133. 10. Blitzer MC, Shapira E: A purified serum glycopeptide from controls and cystic fibrosis patients. I. Comparison of their mucociliary activity on rabbit tracheal explants. Pediatr Res 16:203, 1982. 11. Carson SD, Bowman BH: Cystic fibrosis. I. Fractionation of the mucociliary inhibitor from plasma. Pediatr Res 16:13, 1982. 12. McNeeley MC, Awasthi YC, Barnett DR, et al: Cystic fibrosis. II. The urinary mucociliary inhibitor. Pediatr Res 16:21, 1982. 13. Tummler B, Seger E, Riordan JR: Systematic study of the hydrolysis of 4-methylumbetliferyloguanidinobenzoate in plasma from patients with cystic fibrosis and controls. Clin Chem Acta 125:219, 1982. 14. Branchini BR, Salituro GM, Rosenstein BJ, Bruns WT: 4-methylumbelliferylguanidinobenzoate reactive plasma "protease" in cystic fibrosis is albumin. Lancet 1:618, 1982. 15. Wallevik K: SS-interchanged and oxidized isomers of bovine serum albumin separated by isoelectric focusing. Biochim Biophys Acta 420:42, 1976.
Carrier detection in cystic fibrosis
917
16. Grataroli R, Guy-Crotte O, Galabert C, Figarella C: Detection of the cystic fibrosis protein by isoeleetric focusing of serum and plasma9 Pediatr Res 18:130, 19849 17. Nevin GB, Nevin NC, Redmond AO, et al: Detection of cystic fibrosis homozygotes and heterozygotes by serum isoelectrofocusing. Human Genet 56"387, 1981. 18. Scholey J, Applegarth DA, Davidson AGF, Wong LTK: Detection of cystic fibrosis protein by electrofocusing. [Letter] Pediatr Res 12:800, 1978. 19. Hallinan FM, Kenny D, Tempany E: A study of isoelectric focusing in polyacrylamide gels of serum proteins as a cystic fibrosis screening test. Clin Chim Acta 117:103, 1981. 20. Wilson GB: Monospecific antisera, hybridoma antibodies, and immunoassays for cystic fibrosis protein. Lancet 1:313, 1980. 21. Manson JC, Brock DJH: Development of a quantitative immunoassay for the cystic fibrosis gene. Lancet 1:330, 1980. 22. Bullock S, Hayward C, Manson J, et al: Quantitative immunoassays for diagnosis and carrier detection m cystic fibrosis9 Clin Genet 21:336, 1982. 9
, /
,
A. Rauf Qureshi, Ph.D., and Hope H. Punnett, Ph.D. Philadelphia, Pennsylvania
No TEST is presently available for detection of cystic fibrosis carriers. Although many claims for methods of heterozygote detection have been advanced, few have been confirmed. We have evaluated four reputed markers of the CF gene for their reliability: (1) the cystic fibrosis protein detectable in serum by isoelectric focusing at pH 8.5!; (2) the presence of a lectinlike factor in the serum of CF homozygotes and heterozygotes that agglutinates mouse red cells2; (3) the CF-lectin isolated by affinity chromatography on fructose-conjugated sepharose 6B column3; and (4) the presence of reduced levels of MUGB-reactive proteases in the plasma of CF genotypes. 4,5 We evaluate d these methods for reproducibility in hopes that one alone or a combination of two or more might prove suitable for carrier detection. From the Department of Laboratories, Genetics Section, St. Christopher's Hospital for Children, and the Department of Pediatrics, Temple University School of Medicine. Supported in part by Grant A M 26606 from the National Institutes of Health and by the Cystic Fibrosis Foundation. Submitted for publication Aug. 27, 1984; accepted Dec. 4, i984. Reprint requests." A. Rauf Qureshi, Ph.D., Department of Laboratories, Genetics Section, St. Christopher's Hospital for Children, 5th and Lehigh Ave., Philadelphia, PA 19133.
METHODS Venous blood from patients with CF, obligate heterozygotes, and normal healthy adult controls was collected with informed consent. The controls and heter0zygotes were age matched (range 22 to 54 years). The age range of the patients was 1 to 29 years. The clinical diagnos!s of CF was confirmed by sweat electrolyte analyses. All patients were being seen at the CF clinic at St. Christopher's Hospital
See related article, p. 930. CF MU MUGB
Cystic fibrosis 4-Methylumbelliferone 4-Met hylumbelliferylguanidinobenz0ate
for Children. At the time when blood specimens were taken, most of the patients were receiving supplementarY pancreatic enzymes and antibiotics orally. E!eetrofocusing of CF protein. The concentration of IgG in each sampl e was determined by single radial immunodiffusion using commercial kits (Kailestad Laboratories, chaska, Minn.). Isoelectric focusing in thin-layer polyacrylamide gel using the LKB Muttiphor (Model 2117; LKB Produkter,
The Journal of P E D I A T R I C S
913
9 14
Qureshi and Punnett
The Journal of Pediatrics June 1985
1
2=
3
4
5
6
7"
8 ~
9
Fig. 1. Isoelectric focusing pattern of serum proteins obtained in pH 2.5 to 10.0 gradient. Serum volumes containing 300 or 600 ug lgG were applied on 3 mm pads at anode. Samples were prefocused for 30 minutes at 350 V. then focused for 90 minutes at 35 W. Samples 2.4. 6. and 8 from CF homozygotes, samples 1.3.5.7. and 9 from normal controls. Samples 7. 8, and 9 are same as 3, 2, and 1, respectively, except that 600 ~g IgG was used for samples 7 through 9. Bars indicate presence of CF protein bands. Only samples containing 300 ~zg IgG were considered for presence or absence of CF protein bands. Table I. Presence of CF protein in serum, as determined by isoelectric focusing
In Controls CF heterozygotes CF homozygotes
29 33 64
CF protein band(s) Present Doubtful Absent n
%
n
%
n
5 15 42
17 45.5 65.5
4 10 14
14 " 20 30 8 22 8
% 69 24 12.5
Bromma, Sweden) was performed according to the method of Wilson et al. 6 with unpublished modifications (personal communication, 1980). DeteCtion of CF lectin. Lectinlike activity in the serum w a s m e a s u r e d first according to the earlier methods described by Lieberman and co-workers:. 7 and also according to their modified method (personal communication, 1981). Isolation of CF leetin. Affinity chromatography was performed on fructose-Sepharose 6B column according to the described method of Nordstrom et al? MUGB-reaetive proteases. The titer of MUGB-reactive proteases was measured in citrated plasma of C F homozygotes, heterozygote s, and normal controls by previously described methods.,4 5 Nonspecific hydrolysis of M U G B was corrected by carrying out the reaction in the presence and absence of benzamidine, a competitive inhibitor of serine proteases. M U G B reactivity was measured in unactivated plasma 0nly. ~
!soelectric focusing of activated and unactivated plasma was carried out in 5% polyacrylamide gel, with 0.75% Ampholine, pH 5 to 7, and then stained for M U G B reactivity? RESULTS CF protein. Samples were considered positive for C F protein if a single or a doublet band or a sharp frontal edge was detectable in the most cathodal centimeter of the gel (pI 8.5 to 8.7) (Fig. 1). The C F protein was Present in 46% of heterozygotes and 66% of homozygotes for the C F gene and in 17% of normal controls (Table I). CF leetln. The test was performed in 29 patients with CF, five obligate heterozygotes, and 13 normal controls. Initial results were encouraging. Lectinlike activity was found i n t h e sera from 27% of C F homozygotes and 45% o f heterozygotes, but was present in only 4% of control sera. However, contradictory results were obtained when there was a Change in laboratory personnel, and our earlier results could not be reproduced. Isolation of CF leetin. We were able to desorb a protein peak from the serum of normal controls and C F heterozygotes. A similar peak could also be desorbed from C F homozygotes. The test was performed in four controls, three patients with CF, and four obligate heterozygotes. Lectin was present in all of the sera. Therefore, the lectin peak was not found to be specific for the C F gene (Fig. 2). MUGB-reaetive proteases. Initially, we measured M U G B reactivity on activated and unactivated plasma
Volume 106 Number 6
Carrier detection in cystic fibrosis
2.0
9 15
'
1-6r
o
Ii
N
I I
--
1-2
~
0"8
I, 't
.
iiI
0
~ 0
If'
14
t
1 L
28
42
56 Volume (ml)
112
126
140
^
-
^
154
Fig. 2. Affinity chromatography on fructose-conjugated Sepharose 6B column (0.9 • 15 cm). Column was equilibrated with initial buffer (0.9% saline solution containing 1 mM MgC12, pH 6.0). Four milliliters of serum was diluted with initial buffer and applied to column. Fructose-specific lectin Was eluted (arrow) with initial buffer containing 10 mM EDTA. Effluents were monitored by measuring optical density at 240 nm, using EDTA buffer-as blank..e, Heterozygous serum; ~ normal control. and confirmed (data not shown) the observations of Rao et al. 4 that there was no significant difference in the titers of activated and unactivated plasma. Subsequently, M U G B reactivity was measured in unactivated plasma only. Our results (Table II) show that there was no significant difference in titers or specific activity in plasma from normal controls, heter0zygotes, and homozygotes We did not find any correlation between M U G B reactivity, albumin levels ( X + SD 41.6 _+ 6.4 mg/ml), and a g e of patients with CF. Our data also showed that low M U G B titers ( M U < 200 n m o l / m l plasma) were present in 17% of normal controls, 33% of heterozygotes, and 31% of homozygotes. There was an extensive overlap among the three populations. We were also unable to confirm the differences in the number of bands obtained after isofocusing and staining with M U G B in the plasma of C F genotypes and controls. DISCUSSION The primary goal of this study was the evaluation of published tests purporting to distinguish unaffected individuals who are heterozygous for the C F gene from those who are homozygous normal or who are affected (homozygous for the C F gene) and to see if one method alone or a combination of two or more might prove suitable for carrier detection. Inasmuch as only one method showed some promise, the second part of the study was not done. C F lectin is reported to occur in both C F homozygotes and heterozygotes. The agglutinating activity in C F serum is inhibited by fructose, fucose, and antibiotics. The
Table II. Plasma M U G B activity
(nmol MU/ml)
Specific activity (nmol MU/mg total protein)
232 _+ 38 174 to 285
3.7 _+ 0.5 3.1 to 4.5
225 _+ 44 146 to 293
3.6 _+ 0.7 2.5 to 4.8
214 _+ 46 115 to 304
3.3 -+ 0.8 1.9to5.0
Titer
n Normal controls Mean _+ SD Range Heterozygotes Mean _+ SD Range Cystic fibrosis Mean_+ SD Range
12
24
23
non-CF activity found in some controls is inhibited by mannose. 2 The test as described is simple. However, it is very subjective because there is no sharp end point in scoring agglutinatio n . We have found it difficult to obtain consistent results in successive tests using the same specimen. We did a blind study using specimens collected and coded by the Pulmonary Department, St. Christopher's Hospital for Children. A change of personnel in the laboratory led to contradictory results, reinforcing our concerns about subjectivity. This test was found to be unreliable and unsuitable for screening purposes. Using affinitY column chromatography on fructoseSepharose 6B, we have found the fructose-specific lectin to be present in all" sera tested, including that of normal Controls and C F l~eterozygotes and homozygotes. This finding is contrary to the observations o(]'qordstrom et al., 3 who found that the leetin peak was absent from the sera of controls. However, the authors also pointed out that the
9 16
Qureshi and Punnett
lectin peak cross-reacted with the anti-human serum, suggesting that the lectin was not unique for the CF gene. Lieberman and Kaneshiro9 found a low molecular weight (1000 daltons) lectin in the serum of controls and CF heterozygotes and homozygotes. Only lectin from patients with CF and carriers showed any hemagglutination activity. Similar findings have also been made on CF factors ~~ that are ubiquitous in serum and urine but possess biologic activity only when isolated from CF genotypes. Reduced levels of MUGB-reactive proteases in CF homozygotes and heterozygotes were reported by WalshPlatt et al. ~ We did not find significant differences in the titers of MUGB-reactive proteases in the plasma of CF homozygotes, CF heterozygotes, and controls. However, we did find low titers (MU <200 nmol/ml) in approxi 7 mate!y 33% of CF homozygotes and heterozygotes and in 17% of normal controls. The extensive overlap among the three groups renders this test unsuitable for carrier detection. MUGB has been shown 13 not to meet the requirements of an active site titrant under the conditions described for this assay. Although we did not find any correlation between the levels of albumin and the MUGB reactivity in the plasma of patients with CF, others 14 have found plasma albumin, rather than plasma proteases, to be responsible for MUGB reactivity, because purified albumin was shown to possess MUGB hydrolysing activity. However, the possibility of MUGB-reactive material (e.g., a polypeptide produced as a result of catabolism of arginine esterase in the circulation) being tightly bound to albumin cannot be ruled out. Support for this concept was provided by the isofocusing pattern of MUGB-reactive material in the plasma, inasmuch as identical patterns were obtained in all three populations. This is consistent with the fact that albumin also focuses in the same pI region (5.2 to 5.5) 15 as MUGB-stainable proteases (pI 5.1 to 5.4). 8 Therefore, the determination ~of MUGB reactivity or isofocusing of MUGB-stainable bands for the purpose of carrier detection is not recommended unless--and until--the putative enzyme is identified and characterized: Isoelectric focusing in polyacrylamide gel of serum proteins has been claimed to be a useful method ~for carrier detection. We have confirmed the presence of CF protein in the sera of some individuals heterozygous or homozygous for the CF gene. We have found the CF protein in approximately 46% of heterozygotes and 66% of homozygotes for the CF gene. This is a lower figure than that reported by WilsonI and by others. ~6.t7 Findings similar to ours have also been reported. 18,~9 The frequency of CF protein in our control subjects was 17%, which is higher than the expected carrier frequency in white populations. Although the isofocusing technique is straightforward,
The Journal o f Pediatrics June 1985
requiring no unusual-steps, the bands are often equivocal. They are not always detectable, perhaps because of low levels and lack of sensitivity of the isofocusing techniques. Morever, the method requires some degree of subjectivity in evaluation of results, and is therefore unreliable and unsuitable for population screening, although it may have a role in family studies. The CFP region of polyacrylamide gel has been used to raise antibody in mice2~and in guinea pigs. 2~ Both antisera were reported to distinguish quantitatively between CF homozygotes, heterozygotes, and controls. Bullock et al? 2 failed to produce specific antibody in any species other than guinea pig despite considerable efforts. We have attempted to raise antibody in guinea pigs according to the methods of Bullock, but have not been successful. We conclude that all four methods tested are unsuitable for detection of CF carriers. At present there is no known definite marker for the CF gene that can be used for screening: We thank Lucinda DeMarco and Gregory Higbee for technical assisance; and Dr. Judy Palmer, Dr. Daniel Schidlow, and the staff of the Cystic Fibrosis Center for collecting blood samples. Our special thanks to the patients with CF and 9 parents who donated blood for this study. REFERENCES
1. Wilson GB, Fudenberg HH: Studies on cystic fibrosis using isoelectric focusing. I. An assay for detection of cystic fibrosis homozygotes and heterozygote carriers from serum. Pediatr Res 9:635, !975. 2. Lieberman J, Costea NV, Jakulis VJ, Kaneshiro W: Detection of a lectin in the blood of cystic fibrosis homozygotes and heterozygotes. Trans Assoc Am Physicians 92:121, 1979. 3. Nordstrom OV, Blomfield J, Brown JM: Isolation and characteristics of a fructose specific lectin from cystic fibrosis homozygote and heterozygOte serum. In Sturgess JM, editor: Perspectives in cystic fibrosis. Proceedings of the Eighth International Congress on Cystic Fibrosis, Toronto, Ontario Canada, 1980, p 169. 4. Rao GJS, Walsh-Platt M, Nadler HL: Reaction of 4methylumbelliferylguanidinobenzoatewith proteases in plasma of patients with cystic fibrosis. Enzyme 23:314, 1978. 5. Walsh-Platt M, Rao GJS, Nadler HL: Protease deficiencyin plasma of patients with cystic fibrosis. Enzyme 24:224, 1979. 6. WilsonGB, Arnaud P, Fudenberg HH: Improved method for detection of cystic fibrosis protein in serum using the LKB Multiphor electrofocusing apparatus. Pediatr Res 11:986, 1977. 7. Lieberman J, Kaneshire W, Costea N: Characteristics of a new screening test for detecting the cystic fibrosisgene: Assay of a serum lectin. In Sturgess JM, editor: Perspectives in cystic fibrosis. Proceedings of the Eighth International Congress on Cystic Fibrosis, Toronto, Ontario, Canada, 1980, p 308. 8. Rao GJS, Nadler HL: Arginine esterase in cystic fibrosis of the pancreas. Pediatr Res 8:684, 1974.
Volume 106 Number 6
9. Lieberman J, Kaneshiro W: Detection of a low molecular weight CF-lectin cofactor in serum by G-25 gel filtration. CF Club Abst, 1981, p 133. 10. Blitzer MC, Shapira E: A purified serum glycopeptide from controls and cystic fibrosis patients. I. Comparison of their mucociliary activity on rabbit tracheal explants. Pediatr Res 16:203, 1982. 11. Carson SD, Bowman BH: Cystic fibrosis. I. Fractionation of the mucociliary inhibitor from plasma. Pediatr Res 16:13, 1982. 12. McNeeley MC, Awasthi YC, Barnett DR, et al: Cystic fibrosis. II. The urinary mucociliary inhibitor. Pediatr Res 16:21, 1982. 13. Tummler B, Seger E, Riordan JR: Systematic study of the hydrolysis of 4-methylumbetliferyloguanidinobenzoate in plasma from patients with cystic fibrosis and controls. Clin Chem Acta 125:219, 1982. 14. Branchini BR, Salituro GM, Rosenstein BJ, Bruns WT: 4-methylumbelliferylguanidinobenzoate reactive plasma "protease" in cystic fibrosis is albumin. Lancet 1:618, 1982. 15. Wallevik K: SS-interchanged and oxidized isomers of bovine serum albumin separated by isoelectric focusing. Biochim Biophys Acta 420:42, 1976.
Carrier detection in cystic fibrosis
917
16. Grataroli R, Guy-Crotte O, Galabert C, Figarella C: Detection of the cystic fibrosis protein by isoeleetric focusing of serum and plasma9 Pediatr Res 18:130, 19849 17. Nevin GB, Nevin NC, Redmond AO, et al: Detection of cystic fibrosis homozygotes and heterozygotes by serum isoelectrofocusing. Human Genet 56"387, 1981. 18. Scholey J, Applegarth DA, Davidson AGF, Wong LTK: Detection of cystic fibrosis protein by electrofocusing. [Letter] Pediatr Res 12:800, 1978. 19. Hallinan FM, Kenny D, Tempany E: A study of isoelectric focusing in polyacrylamide gels of serum proteins as a cystic fibrosis screening test. Clin Chim Acta 117:103, 1981. 20. Wilson GB: Monospecific antisera, hybridoma antibodies, and immunoassays for cystic fibrosis protein. Lancet 1:313, 1980. 21. Manson JC, Brock DJH: Development of a quantitative immunoassay for the cystic fibrosis gene. Lancet 1:330, 1980. 22. Bullock S, Hayward C, Manson J, et al: Quantitative immunoassays for diagnosis and carrier detection m cystic fibrosis9 Clin Genet 21:336, 1982. 9
, /
,