Partial Characterization, Properties, and Clinical ... - Clinical Chemistry

CLIN.CHEM.29/2, 272-278 (1983)

Partial Characterization, Properties, and Clinical Significance of a Lactate Dehydrogenase-Immunoglobulin AK Complex in Serum R. N. M.

Weljers,1 J. Mulder,2 and H. Kruijswljk1

The existence of a lactate dehydrogenase-immunoglobulin A,, complex was demonstrated as a marked increase of

lactate dehydrogenase activity at the lactate dehydrogenase3 band in the agar-agarose gel-electrophoresis pattern of sera from seven patients from an unselected group of 21 800 patients. The complex was isolated, in an almost pure form, by gel filtration and affinity chromatography, from the serum of a patient with circulating hepatitis B surface antigen. The complex had a relative molecular mass (Mr) of approximately 445 000 as determined by gel filtration. Electrophoresis in sodium dodecyl sulfate in the absence of reducing agents showed the presence of two subunit bands with Mr 170 000 and 34 000. We therefore propose that the native complex consists of one monomeric immunoglobulin A,, linked to two tetrameric lactate dehydrogenase molecules. The enzymic activity of the complex is probably from lactate dehydrogenase isoenzyme-5.

AdditIonal -

Keyphrases: hepatitis B determination of relative molecular mass

-

isoenzymes

Several studies on the isoenzyme pattern of lactate dehydrogenase (LD; L-lactate:NAD oxidoreductase, EC 1.1.1.27) in serum from patients with different diseases and from clinically asymptomatic individuals indicate the presence of LD isoenzymes linked to immunoglobulin A of the kappa or lambda type [IgA,,, IgA; (1-11)1. Regardless of the electrophoretic techniques used-i.e., agarose gel electrophoresis(1), agar gel electrophoresis (2-4,6-8, 10, 11), or cellulose acetate electrophoresis (5, 9)-the complexes are associated

with abnormal

LI) isoenzyme patterns characterized by the absence of the LD-2 isoenzyme (2-6, 8, 9); the presence of one or more additional LD isoenzymes (2-9); an

alteration of the electrophoretic mobility of the LD isoenzymes (1); the presence of a broad, diffuse zone of LD activity (4, 11); or a marked increase of activity at the LD-3 band (3, 4, 8, 9).

LD isoenzyme-IgA complexes from patients with different diseases and from clinically asymptomatic individuals have also been studied with use of gel-filtration chromatography (1-4, 7-9). These studies demonstrate not only LD isoenzymes of small molecular size but also LD isoenzyme-IgA complexes of large molecular size. From these studies one also may conclude that one LD isoenzylne (3, 8), two LD isoenzymes (3, 4, 8, 9), or all five U) isoenzymes (1, 8) are involved

in the formation

of the complex.

‘Department of Clinical Chemistry, Onze Lieve Vrouwe Gasthuis, le Oosterparkstraat 179, 1091 HA Amsterdam, The Netherlands. 2 Immunology, Laboratorium voor de Volksgezondheid, Jelsumerstraat 6, 8917 EN Leeuwarden, The Netherlands. Nonstandard abbreviations: LD, lactate dehydrogenase; IgA, IgG, 1gM, immunoglobulins A, G, and M, respectively; HBs, hepatitis B surface; and INT, 2-(p-iodophenyl)-3-(p-nitrophenyl)-5-phenyl211-tetrazolium chloride. Received July 14, 1982; accepted Oct. 26, 1982. 272

CLINICAL CHEMISTRY, Vol. 29, No. 2, 1983

The published literature through 1981 records six patients whose serum LD isoenzyme pattern showed a marked increase of activity at the LD-3 band, due to a LD-IgA,, complex, but with all five electrophoretically normal LD isoenzyme bands also present (1,2). This group of patients is characterized as follows: a large molecular form of LD, having the electrophoretic mobility of the LD-3 isoenzyme in agar gel, circulates in patient’s serum. The LD activity in serum is normal to moderately increased. One of the serum proteins is a LD-IgA complex, which is not dissociated in the presence of added NAD. The light chain involved in the complex formation is of the kappa type. The LD activity in the LD-IgA complex is localized only on a small part of the serum IgA precipitation line, cathodic to the application point. The concentrations of IgA, IgG, and 1gM in serum are within the reference intervals (6). Finally, in these patients the relationship between the diagnosis and the existence of the LD-IgA complex is uncertain. In serum from a patient with hepatitis B surface (HBs) antigens we observed, by use of agar-agarose gel LD isoenzyme electrophoresis, a marked increase of LD activity at the LD-3 band. Immunoelectrophoresis followed by assay of LD activity demonstrated the presence of an LD-IgA,, complex.

We report

and its composition,

here

our purification

Also, we examine

of this

the clinical

patients having serum with the above-mentioned and evaluate the complex in terms of its clinical

and its diagnostic

complex,

data for features chemistry

significance.

Materials and Methods Subjects We studied seven subjects, selected because their serum showed markedly increased activity at the LD-3 band in the agar-agarose LD isoenzyme pattern. The five men and two women, ages 20 to 64 years (median: 35 years), were admitted to the Departments of Internal Medicine and Cardiology of the Onze Lieve Vrouwe Gasthuis. The subject from whose serum we purified the LD-1gA complex was a 20-year-old woman with HBs antigens in her serum.

Chemicals Routine chemicals of analytical grade were supplied (except where noted) by Merck, Darmstadt, F.R.G. The biochemicals were generally of the highest purity available and were obtained from the following sources: LC-Partigen-IgA radial immunodiffusion plates and human protein standard serum LC-V (Behring Diagnostics, Marburg, F.R.G.); bovine albumin standard and Coomassie Brilliant Blue G-250 (BioHad Laboratories, Richmond, CA 94804); NAD and NADH (Boehringer Mannheim GmbH, Mannheim, F.R.G.); phenazine mbthosulfate (PMS; Calbiochem-Behring Co., La Jolla, CA 92037); monospecific rabbit antisera (anti-y-, a-, /.L-, K-, and X-chains; Dako, Copenhagen, Denmark, or Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands); agar Noble (Difco Laboratories, Detroit, MI 481412); agarose (Litex, Denmark); Ultrogel AcA 22 and Ultrogel AcA 44 (LKB, Stock-

hoim, Sweden); 2-(p-iodophenyl)-3-(p-nitrophenyl)-5-phenyl2F1-tetrazolium chloride (INT; Merck, Darmstadt, F.R.G.); 5’-AMP-Sepharose 4B [.N6-(6-aminohexyl)-5’-AMP bound to Sepharose 4B], carbonic anhydrase (EC 4.2.1.1), ovalbumiii, albumin, phosphorylase kinase (EC 2.7.1.38), aldolase (EC 4.1.2.13), catalase (EC L11.1.6), ferritin, and thyroglobulun (Pharmacia, Uppsala, Sweden); and a-chymotrypsinogen, IgG, and transferrun (Sigma Chemical Co., St. Louis, MO 63178). All solutions were prepared in de-ionized water. Procedures Prelii’ninary

30#{176}C with

assays.

the

LD activity was assayed at analysis system (Boehringer Bromma, Sweden) according the by the Netherlands Society for ClimTotal

“PRISMA”

Mannheim-Clinicon, method recommended cal Chemistry (12). The 95th percentile for serum LD activity in normal persons by this metHod in our hands is 275 UIL. Serum immunoglobulins were quantified by single radial imrnunodiffusion (13). Protein concentrations were measured by the biuret assay (14) or by the Coomassie Brilliant Blue G-250 assay (15). To determine whether free IgA,, was associating with U), we incubated, at 37 #{176}C for 15 mm, test mixtures of a patient’s serum mixed with an equal volume of a control serum and a patient’s serum mixed with an equal volume of a serum having a relatively high LD-5 content. These were then subjected to agar-agarose gel electrophoresis. Tests for hepatitis antigens and antibodies were performed by radioimmunoassay (Abbott Laboratories, North Chicago, IL 60064) at the Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam,

The Netherlands. Ultrafiltration. Samples were concentrated with ultraiiltration equipment (Model 402 and Model 8MC) and ultrafiltration membranes [PM-b (10 000 Mr cutoff) and PM-30 (30 000 M, cutoff)] from Amicon Corp., Lexington, MA 02173. Electrophoresis. For agar-agarose (7/3, by wt; 9.8 g/L total) gel electrophoresis we used gels containing 36 mg of disodium EDTA and 196 mg of human albumin per liter of barbital buffer (50 mmol/L, pH 8.4), essentially according to Wieme (16), with the following modifications. On each half of the 76 x 26 x 2.5 mm slides we applied two 2-giL serum samples 30 mm apart, then electrophoresed at 5 #{176}C and 25 mA per slide (constant current), for 15 mm. To stain for U) activity, we covered the agar-agarose gel with a 2.5-mm thickness of 8 gIL agar gel in barbital buffer (54 mmol/L, pH 8.4) containing 90 mmol of sodium lactate, 40 mg of disodiurn EDTA, 900 mg of NAD, 240mg of INT, and 24 mg of phenazine methosulfate per liter and incubated for 2 h at

37#{176}C.

Polyacrylainide disc gel electrophoresis was done by the QDA” specific reagent-kit instructions (no. 3510, revised 10/74; Miles Labs., Elkhart, IN 46514). Detection of LD activity was based on its conversion of INT, for whici we used the LD isoenzyme substrate set of Corning, Palo Alto, CA 94306. A duplicate gel loaded with the same sample was stained for proteins with Coomassie Brilliant Blue G-250. After immunoelectrophoresis as described by Scheidegger (17), U) activity was detected according to Brendel et al. (18). A duplicate gel loaded with the same sample was stained for proteins with Coomassie Brilliant Blue G-250. “Quick-Disk

Determination of relative molecular mass. Gel electrophoresis in sodium dodecyl sulfate was done according to Weber and Osborn (19). Before applying the sample to the gels, we

boiled them

for 2 mm in 10 g/L sodium

dodecyl sulfate 2-mercaptoethanol. We used a log plot of the Mr values of reference proteins vs their respective gel migration to establish the Mr of the complex. To determine the Mr of the purified protein under native conditions, we applied it to an Ultrogel AcA 22 column (1.4 x 100 cm) equilibrated with Tris HC1 buffer (100 mmoIJL, pH 8.0) containing, per liter, 500 mmol of NaC1 and 0.2 g of NaN3 (buffer A). Fractions of 2.35 mL were collected at a flow rate of 4.7 mL/h. Purification. Blood (100 mL total) from the median cubital vein was collected into 10 separate glass tubes and allowed to clot by standing at 37 #{176}C for 30 mm. The clot was removed by centrifugation (3360 x g, 15 mm). After recentrifuging (3360 x g, 15 mm) the 50-mL of combined supernate, we added 02 g of NaN3 per liter of serum (Pool X0) and passed 8.0-mL samples of this pool through an Ultrogel AcA 44 column (5.0 x 90 cm) with buffer A. Fractions of the effluent were assayed for LI) activity, and those corresponding to the two peaks with LD activity were pooled separately (Pool X1 and Pool Y1) and were respectively concentrated to about 30 mL each by ultrafiltration through a PM-30 membrane. We then assayed the concentrated Pools X1 and Y1 for U) activity and examined them by electrophoresis and immunoelectrophoresis. Because Pool Y1 showed no evidence of the LD-IgA complex, we discarded it from further study. To remove most of the Tris HC1 buffer, we diluted the concentrated Pool X1 fivefold with phosphate buffer (100 mmol/L, pH 7.0), then reconcentrated it to about 40 mL. After centrifi.igation at 22 000 x g for 60 mm, we applied the Pool X1 supernate to a 26 x 70 nun column containing 10 g of 5’-AMP-Sepharose 4B gel, perfusing the sample through the gel according the method of Pettit et a!. (20). After assaying the eluted fractions for U) activity, we combined those that had U) activity (which eluted with 0.4 mmol/L NADH solution) into Pool X2. Pool X2 was concentrated to about 3 mL by ultrafiltration with a PM-b membrane and analyzed by electrophoresis. To remove NADH from the concentrated pool X2, we passed the pool through an Ultrogel AcA 44 column (1.4 x 100 cm) with buffer A at a flow rate of 4.7 mL/h, collecting 2.35-mL fractions. The fractions were assayed for U) activity and those with U) activity were combined into Pool X3. Pool X3 was concentrated to about 2.7 mL by ultrafiltration through solution

without

a PM-b

membrane.

Results Purification of the LD-lgA,, Complex Given

the presence of an abnormal agar-agarose U) pattern, i.e., a marked increase of U) activity at the U)-3 band (Figure IA), in the serum of a 20-year-old woman with HBs antigens, we assumed the presence of a LD-immunoglobulin complex. Immunoelectrophoresis, followed by assay for U) activity, demonstrated the presence of a LD-IgA,, complex (Figure 2). The complex was purified about 10000-fold from the serum to apparent homogeneity (Table 1). When Pool X0 (8 mL per run, total U) activity = 650 U/L) was applied to an Ultrogel AcA 44 column, the U) activity was eluted in two peaks (Figure 3A). In all, 48 mL of Pool X0 was passed through the AcA 44 column. After concentration, 40 ml of Pool X1 contained 17.6 g of protein and 430 U of U) activity per liter. The calculation of LD activity recovered was based on the LD-IgA, complex containing 65% of the total LI) activity of the original serum sample, as estimated by peakarea measurements (Figure 3A). Agar-agarose gel electrophoresis of concentrated Pool X1 demonstrated LD activity isoenzyme


273

Table 1. PurifIcation of LD-IgA, Vat, mL

Step

Original serum

(Pool X)

Ultrogel AcA 44 chromatography, followed by concentration (Pool X1) 5’-AMP.--Sepharose 46 chromatography, followed by concentration (Pool X2) Ultrogel AcA 44 chromatography, followed by concentration (Pool X3)

Complex from Human Serum

Total LD-IgA. acty, mU

48 40

20 280 17 200

3

16 800

2.7

5

Total

Spec. acty.,

proteIn, mg

U/g

Recovery,

factor

5.3

3835

705

PurificatIon

1

100 84.8

24.4

4.6

82.8 0.1

700

57 000

28.1

10 775

+

B

A

+ N

1

...,_ant

-r,-#{149}

i-I9A

P

.E

.

7..I

N 2 P

N

.1l

-

.........anti-IgA

3 P

C N _

4

anti- c

P

Fig. 2. lmrnunoelectrophoresis of original serum sample followed by 1,2), or followed by staining for protein with Coomassie Bnltiant Blue G-250 (panels 3, 4) N, normal serum; P, patients serum

assay of LD activity (panels

LD patterns after the first stage of purification on serum LD-lgA,, complex Fig. 1. Agar-agarose

Original serum sample; B, pooled fractions of low-M, ID activity obtained after Ultrogel AcA 44 column chromatography; C, pooled fractions of high-M, LD activity obtained after Ultrogel AcA 44 column chromatography. U) activity applied (at points indicated by arrow) to gels A, B, and Cwas 1250,375, and 475 MU,respectively

only at the U)-3 band (Figure 1C). Immunoelectrophoresis, followed by assay for U) activity, demonstrated the pres-

ence of the LD-IgA,, complex. Agar-agarose electrophoresis of the concentrated Pool Y1 material showed a normal distribution of U) activity among the five characteristic U) isoenzymes (Figure 1B); immunoelectrophoresis followed by assay of U) activity demonstrated the complete absence of the LD-IgA,, complex. Therefore, we no longer studied this pool.

The concentrated affinity 274

chromatography

Pool

X1 was diluted buffer, applied

fivefold with to the


the

5’-AMP-

Sepharose 4B column, and then eluted with the washes shown in Figure 3B. U) activity was eluted as a single peak by the 0.4 mmol/L NADH wash. The fractions containing U) activity were pooled (Pool X2; 40 mL) and concentrated to 3 mL; total U) activity was 5600 U/L. To remove the NADH, the concentrated Pool X2 was passed through an Ultrogel AcA 44 column (Figure 3C). After concentration of the fractions having LD activity (42 mL concentrated to 2.7 mL, Pool X3), total U) activity was 2110 U/L. Concentrated Pool X3 contained 100 zg of protein, including 22 pg of IgA protein; immunoelectrophoresis followed by assay for LD activity indicated the presence of the LD-IgA,, complex. Electrophoresis of 3.7 pg of concentrated Pool X3 on standard polyacrylainide disc gel followed by assay for U) activity demonstrated a single, densely staining band with U) activity (Figure 4A). A second gel, stained with Coomassie Brilliant Blue G-250, revealed a single band with the same Rf as the densely staining band with

U) activity

(Figure

4B).

Characterization of the LD-lgA,, Complex We estimated

the relative

molecular

mass

of the purified

15

A

A

60

-. D

B

S

40 8

a

05

-J

20

0

B

0.

1000

800 600 0

a

#{163}00

#{163}00

200

0 C

ml 0

1200

-J

800

S

8

8 600 a

.4

4

40)

-J

20)

0

x3

Fig. 3. LO activity (#{149} and broken line) and absorbance (solid line) of 125-pt aliquots of fractions eluted from the patient’sserum A, elution pattern obtained with Uftrogel AcA 44 filtration; B, elution pattern obtainedby affinityctwomategraphyon 5’-AMP--Sepharose46; C, elutionpattern obtainedwith UltrogelAcA 44 filtratIon.For details,see Materialsand Methods native LD-IgA,6 complex from its elution properties in gel filtration with Ultrogel AcA 22. The partition coefficient of the LD-IgA,, complex chromatographed as described above correlates with an approximate Mr of 445 000 (SD - 17000; Figure 5). We estimated the Mr of each subunit of the LD-

Fig. 4. Disc gel electrophoresis of purified LD-IgA,. complex followedby for ID activity (A) or by staining for protein with Coomassie BrilliantBlue G-250 ( assay

IgA complex by electrophoresis of concentrated Pool X3 under non-reducing circumstances on polyacrylamide gels (100 g/L) containing sodium dodecyl sulfate. Migration distances of two subunits, i.e., monomeric IgA and monomeric U), indicated Mr’5 of 170 000 and 34000, respectively, by comparison with proteins of known M. (Figure 6). Ii0l

(x

10”)

(5 UI

80 70 60

0

200

,

A

1: E50

cotolose 0 0I

aldotose 0.1 IA

ii

0

I

I

I

I

I

I

0.3

0.4

0.5

0.6

0.7

0.8

K0

Fig. 5. Determinationof the M of purified native LD-IgA,, complex (A) by AcA 22 chromatography Fordetails,see Materials arid Methods. The M,’sof the comparisonproteinsare: rabbit muscle aldolase, 158 000; bovine liver catalase,232 000; horse spleen ferritin,440 000; and bovinethyroidthyroglobulin,669 000

0.2

0.3 0.4 0.5 0.6 Relative mobility

0.7

Fig. 6. Estimation of the Mr’Sof the subunits (A, complex by sodium dodecyl sulfate polyacrylamide

0.8

0.9

of purified LD-lg gel electrophoresis

in the absence of reducing agents Fordetails,see Materialsarid Methods.The M,s of the comparisonproteinsare: IgG, 160 000; phosphOrylase kinase (phosphonjlase B), 94000; transfemn, 76000; bovineserum albumin. 68000; ovalbtimin, 43000; carbonic anhydrase, 30 000; and a-thyrnotlypsiflogen, 25700 CLINICAL CHEMISTRY, Vol. 29, No. 2, 1983

275

Table 2. Cilnicai and Clinical Chemical Data for Individuals Whose Serum Showed the Presence of an LD-igA Complex Mean serum

Subject

A B C

DiagnosIs

U/Lb

23 f, 28

HB antigen present Apparently healthy

370 365

Myopathy

280 410 375

LD acty.,

,

,

D E F

50

d, 39 9, 35 2, 20 , 64

G M

Sex and age, yr

Apparently healthy Apparently healthy HB antigen present

625

Months elapsed after first

LD-3 acty., % of total serum LO

observation

RelatIve

acty.#{176}

0.34

0.0260

0.35

n.d.

0.34 0.35

0.0269

25 12 41

44 58 33

60 29

48 51

9

61

0.35 0.35

68

0.34

Cardiac asthma 600 7 at the LD-3 band after electrophoresis on agar-agarose gel.

Large-

mobilityd

n.d. 0.0271 0.0265 n.d.

defined In text, with increased LD activity

#{176}Reference value: 95th percentile: 275 U/L. 0Referencevalue(% of total serum LD activity), mean (±SD): LD-3 = 19.1 (1.6)%. d If the distances migrated by the LD-lgA. complex and the most basic lgA are 1 and 12, respectively,the relative mobility R12 nd. not done. Therefore, we conclude that one molecule of monomeric kappa type, is linked with two molecules of tetrameric

Clinical

IgA, U).

Observations

H:

S

11/12

1.

During a six-year study the agar-agarose U) isoenzyme patterns for serum of seven subjects from an unselected population of 21 800 outpatients demonstrated a marked increase of U) activity at the U)-3 band in the presence of all five electrophoretically normal U) isoenzyme bands. Table 2 summarizes the clinical data for these subjects. Although subjects C and G suffered from myopathy and cardiac asthma, respectively, the remaining subjects were apparently healthy. Subject B was referred to our outpatient clinic for evaluation of epigastric pain; and A, D, E, and F were referred for evaluation of a long-standing complaint of over-tiredness.

For all of these

subjects the increased total serum LD (Figure 7) and the marked increase of LD activity at band in the agar-agarose pattern (Table 2) were from the beginning and have persisted so far interruption. In all seven the activities of serum phosphatase (EC 3.1.3.1), alanine aminotransferase (EC 2.6.1.2), and y-glutamyltransferase (EC 2.3.2.2) were less than their respective 95th percentile values. Immunoelectrophoresis, followed by assay for LD activity, revealed in all cases the presence of an LD-IgA. complex. As in previous reports (3, 8), in all cases the U) activity was present on a small and constant part of the precipitation line of IgA, cathodic to the application point. The agaragarose serum LD pattern of the patients did not change on addition of as much as lOg of NAD per liter of serum. The U) isoenzyme patterns of the test mixtures containing patient’s serum and control serum, and of the test mixtures activity the U)-3 present without alkaline

containing patient’s serum and serum with a relatively high amount of U)-5 did not demonstrate a relative increase of U) activity at the U)-3 band.

Table 3. Occurrence

Fig. 7. Time-course studies on the serum ID activity of the seven subjects The dotted lines indicatethe 95th percentile of the referenceintervalfor serum LD activity; the aims’s indicate the dates on which the serological tests for the diagnosisof hepatitis were performed To investigate a correlation between an infection with hepatitis virus and the existence of LD-IgA complex (21), we performed serological tests for HBs antigen, hepatitis B core antigen, hepatitis B 0e” antigen, and their respective antibodies (anti-HBs, anti-HBc, and anti-HBe), and for antthepatitis A antibody (see Figure 7); the results are summarized in Table 3. One or more of the test results was

of Hepatitis B Antigens, Anti-Hepatitis B Antibodies, in Sera Having an LD-IgA. Complex

and Anti-Hepatitis

A Antibody

Subject A

HBsAg Anti-HBs

+

Anti-HBc

+

HBeAg Anti-HBe Anti-HA +,

276

B

-

-

D -

E -

F +

G -

+ +

+

antigen/antibody demonstrable;

C -

-,

nd. nd. + n.d. n.d. antigen/antibodynot demonstrable; nd. not done


nd. n.d. +

+

+ +

positive in five cases, whereas in the other two cases neither antigens nor antibodies were demonstrable. For a more detailed study of the LD-IgA complex, we collected, with the subjects’ permission, 40-mL blood samples from A, D, E, and F. For each, the elution proffle from the Ultrogel AcA 44 column (Figure 3A) clearly indicated both a large (peak X) and a small (peak Y) molecular size LD. The agar-agarose LD pattern for peak X material revealed U) activity only at the LD-3 band, whereas the peak Y material had a normal distribution of LD activity among the five characteristic LD isoenzymes (Figure 8; for the results for subject F, see Figure 1). Immunoelectrophoresis followed by assay for LD activity demonstrated that the LD-IgA complex was present only in peak X material and was completely absent from peak Y material. Finally, the serum concentrations of IgG, IgA, and 1gM were all within the reference intervals for normal subjects.

I cv

/3

lii

]It

a

a

/3

P

Fig. 8. Agar-agarose LD patterns after column chromatography of serum LO-IgA,, complex on Ultrogel AcA 44 I, case A; Il. case 0; III, case E. a, /3, and y representoriginal serum sample, fractionwithlow-M,LD activity,and fraction with high-M, LD activity, respectively

Discussion The LD-IgA,, complex (which has the same electrophoretic mobility as the U)-3 isoenzyme) and the low-Mr U) isoenzymes account for all the U) activity present in the original serum specimen, because the yield of U) activity after the first passage through the Ultrogel AcA 44 column was almost 100%. Our results indicate that the 100 g of the final preparation, was almost homogeneous. The estimated Mr of the native complex (approximately 445 000) and of the dissociated subunits (170000 and 34000, representing monomeric IgA and monomeric U), respectively) rules out the presence of secretory IgA (Mr = 390 000) (22) or of polymers of IgA. Although one can conclude from these data that the complex contains one molecule of monomeric IgA associated with two molecules of tetrameric U), the estimated protein content ratio of IgA to IgA-2U) based on our results is 0.22 instead of the theoretical value of 0.38. Two factors may be responsible for this discrepancy. Firstly, we used a lyophilized standardized protein preparation of bovine plasma albumin to estimate the protein content of preparation X3, which contained both X3,

IgA and U); however, from Wilgenburg et al. (23) it appears that the sensitivity of Ccomassie Brilliant Blue G-250 differs significantly for different proteins. Secondly, the immunological estimation by single radial immunodiffusion of the IgA content of preparation X3 is uncertain because some of the antigenic determinants of IgA may be shielded by the presence of U) (24). Relative molecular masses of various complexes in which an Ig molecule is associated with an enzyme molecule have been reported on several occasions: amylase-IgA (25), creatine kinase-IgG (26, 27), creatine kinase-IgA and alkaline phosphatase-IgG (29). To our knowledge, however, no data on Mr values of immunoglobulin-LD complexes have been reported. We assigned 445 000 ± 17000 for the Mr of a native complex consisting of one molecule IgA,, associated with two U) molecules. Because IgA contains two antigen-binding sites per molecule, an antibody-antigen nature of the binding between IgA and LD in the LDIgA,, complex is strongly suggested. These results extend the previous observations (25-29) of macro-enzymes involving an immunoglobulin molecule associated with two enzyme molecules. By electrophoretic separation of LD isoenzymes on a medium exerting little or no retarding effect related to molecular size, i.e., agar or agar-agarose gel, one may obtain a pattern characterized by a marked increase of U) activity at the LD-3 band in the presence of all five U) isoenzyme bands. The increase of LD activity at the U)-3 band may even exceed the activity of the LD-2 band. To our knowledge, this well-defined LD pattern does not correlate with any particular disease. However, the data for our seven cases that had this characteristic LD pattern confirm the data of a case reported by Nagamine (3), and of five cases reported by Biewenga and Feltkamp (8). Therefore, we combined the results for the 13 cases and sununarize the features of this group as follows: Of all 13 subjects (eight men and five women, ages 20 to 73 years), five were younger than 29, indicating that the onset of the phenomenon might occur at an early age. Four of the 13 were apparently healthy, apart from a long-standing complaint of overtiredness, and the diagnosis of the remaining nine failed to explain the biochemical abnormalities. Thus, the pathogenic mechanism resulting in the above U) pattern is so far unknown. The p1 of the LD-IgA,, complex was, within narrow limits, constant; the LD activity was localized on a small part of the serum IgA precipitation line, cathodic to the application point. Furthermore, in all seven of our subjects the relative mobility of the LD-IgA,, complex was constant (Table 2). The relative molecular mass also was constant, as reflected by the constant partition coefficients we estimated by Ultrogel AcA 44 chromatography (Table 2). On diagnostic grounds, in four of our seven subjects we could expect the U)-5 activity to be increased: in A and F because of the presence of hepatitis B antigen (30,31), in C because of myopathy (32), and in G because of hepatic lesion associated with cardiac asthma (33). Although we cannot at this time provide experimental evidence concerning the nature of the LD isoenzyme of the LD-IgA,, complex, we believe the clinical data on these four patients favor a complex of IgA,, with the LD-5 isoenzyme. The results of the test-mixture experiments led to the conclusion that either the concentration of free IgA,, associating with U) was below the detection limit of our procedures or the free IgA,, was unable to bind the LD from other patients. Our findings show that the serum LD-IgA,, complex was present during a period ranging from seven to 60 months (mean = 26 months). This result contrasts with other reports on the biological half-lives of U) [52 h (34)1 and IgA [six to seven (28),

CLINICALCHEMISTRY, Vol. 29, No. 2,

1983

277

days (35, 36)1. Given the results of this study and of a previous study (21), the initiating trigger mechanism may well be an infection with a virus, probably hepatitis B. In our study, three subjects showed serological evidence of infection with hepatitis B. Hepatitis B antigens or antibodies were undetectable in the other four, either because there were none present or because of low titers as a result of the time elapsed between the onset of the infection and the performance of the tests. We expect that further studies will lead to a more detailed insight into the binding site on the IgA,, molecule for the LD molecule and into the pathogenic mechanism that produces this well-defined LD isoenzyme pattern. We thank Drs. R. Bakker, J. Boomgaard, M. R Esseveld, and C. L. van Schaik for permission to include their patients in the study; Ms. H. van Meyel, Ms. J. Modderman, and Mr. H. van der Mejjden for technical assistance; and Ms. M. de Vries for secretarial assistance.

References 1. Ganrot P0. Lupoid cirrhosis with serum lactic acid dehydroglinked to an ‘A immunoglobulin. Experien/.ia 23, 593 (1967). 2. Biewenga J, Thys LG. Lactate dehydrogenase isoenzyme(s) linked to IgA immunoglobulin in a patient with a myocardial infarction. Clin Chim Ada 27, 293-299 (1970). 3. Naganune M. Lactate dehydrogenase isoenzyxnes linked to immunoglobulin A in two cases. Clin Chim Acta 36, 139-144 enase

(1972).

15. Bradford M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 72, 248-254 (1976). 16. Wieme RJ. Studies on Agar Gel Electrophoresi.s. Arscia, Brus-

sels, 1959, pp 83-96. 17. Scheidegger JJ. Une micro-m#{233}thode de l’immunoelectropho-

r#{232}se. mt Arch Allergy 7, 103-110 (1955). 18. Brendel 8, Mulder J, Verhaar MAT. A rapid and sensitive method for LDH-isoenzyme determination. Int Res Commun System (73-10) 3-1-25. 19. Weber K, Osborn M. The reliability of molecular weight determinations by dodecyl sulfate-polyacrylamide gel electrophoresis. J Biol Chem 244, 4406-4412 (1969).

20. Pettit SM, Nealon DA, Henderson AR. Purification of lactate dehydrogenase isoenzyme-5 from human liver. C/in Chem 27, 8893 (1981). 21. Vyas GN, Peterson DL, Townsend RM, et al. Hepatitis B “e” antigen: An apparent association with lactate dehydrogenase isozyme-5. Science 198, 1068-1070 (1977). 22. Tomasi TB. The gamma A globulins: First line of defense. In Immunobiology, RA Good and DW Fisher, Eds., Sinauer Associates, Inc., Stamford, CT 06905, 1971, pp 76-83. 23. Wilgenburg MGM, Werkman EMA, Gorkom WH, Soons JBJ. Criticism of the use of Coomassie Brilliant Blue G-250 for the quantitative determination of proteins. J C/in Chem C/in Biochem 19, 301-304

(1981).

24. Mulder J, Sloots LCE, Verhaar MAT. Molecular heterogeneity effects of immunoglobulin classes in radial immunodiffusion. J Immunol Methods 1, 211-213 (1972). 25. Harada K, Nakayaina T, Kitamure M, Sugirnoto T. Immunological and electrophoretical approaches C/in Chim Acta 59, 291-299 (1975).

to macroamylase

analysis.

4. Biewenga J. Serum lactate dehydrogenase isoenzymes linked to immunoglobulin A. Clin Chim Acta 40, 407-414 (1972). 5. Markel SO, and Janick SL. Complexing of lactate dehydrogenase isoenzymes with immunoglobulin A of the kappa class. Am J C/in Pat/wi 61, 328-332 (1974). 6. Biewenga J. Thesis, University of Amsterdam, The Netherlands, 1974. 7. Thomas DW, Rosen SW, Kahn CR, et al. Macromolecular lactic acid dehydrogenase. Ann Intern Med 81, 434-439 (1974).

26. Stein W, Bohner J, Eggstein M. Eigenschaften und Bedeutung von Makrokreatinkinasen. J C/in Chem C/in Biochem 17, 739-745

8. Biewenga J, Feltkamp

29. Crofton PM, Smith AF. The properties and clinical significance of some electrophoretically slow forms of alkaline phosphatase. C/in Chim Acta 83, 235-247 (1978).

TRW. LDH-IgA

immunoglobulin

com-

plexes in human serum. C/in Chim Acta 58, 239-249 (1975). 9. Trocha PJ. Lactate dehydrogenase isoenzymes linked to betalipoproteins and immunoglobulin A. C/in Chem 23, 1780-1783 (1977). 10. Biewenga J, van Loghem E. Antigenic analysis of the IgA component of LDH-IgA immunoglobulin complexes. C/in Chim Acta 82, 201-204 (1978). 11. Leroux-Roels GO, Wieme RJ, de Broe ME. Occurrence of enzyme-immunoglobulin complexes in chronic inflammatory bowel disease. J Lab C/in Med 97, 316-321 (1981). 12. The Committee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology,Recommended methods for the determination of four enzymes in blood. Scand J Clin Lab Invest 33, 291-306 (1974). 13. Mancini G, Carbonara AO, Heremans JF. Iminunochemical quantification of antigens by single radial inimunodiffusion. Immunochemistry 2,235-254(1965). 14. Gornall AG, Bardawill CJ, David MM. Determination of serum proteins by means of the biuret reaction. JBioi Chem 177, 751-766 (1949).

278


(1979).

27. Stein W, Bohner J, Krais J, et al. Macro creatine kinase BB: Evidence for specific binding between creatine kinase BB and immunoglobulin G. J C/in Chem C/in Biochem 19, 925-930 (1981). 28. Yuu H, Ishizawa S, Takagi Y, et al. Macro creatine kinase: A study on CK-linked immunoglobulin. C/in Chem 26, 1816-1820 (1980).

30. Bianchi L, Zimmerli-Ning M, Gudat F. Viral hepatitis. In Pathology of the Liver, RNM MacSween, PP Anthony, PJ Scheuer, Eds., Churchill Livingstone, New York NY 10036, 1979, pp 164-

191. 31. Demetriou JA, Drewes PA, Gin JB. Enzymes. In Clinicoj Chemistry, 2nd ed., JU Henry, DC Cannon, JW Winkelman, Eds., Harper and Row, Hagerstown, MD 21740, 1974, pp 818-1001. 32. Wilkinson JH. Diagnostic enzymology. In Scientific Foundations of Clinical Biochemistry, I, DJ Williams, RF Nunn, V Marks, Eds., Heineman, London, 1978, pp 239-249. 33. Bras G, Brandt KH. Vascular disorders. In ref. 30, pp 315-334. 34. Diem K, Lentner C. In Wissenschaflhiche Tabellen, JR Geigy, Ed., Pharma, Basel, 1968, p 582.. 35. Henry JB. Clinical chemistry. In Clinical Diagnosis, 15th ed., I Davidsohn and JB Henry, Eds., Saunders, Philadelphia, PA 19105, 1974, pp 516-664. 36. Nadorp JHSM. Thesis, University of Nijmegen, The Nether-

lands, 1974, pp 87-125.