metallothionein and occupational to cadmium · metallothionein (mt) is a low molecular weight (

British Journal of Industrial Medicine 1983;40:305-313

Metallothionein and occupational exposure tocadmiumF Y FALCK JR,' L J FINE,' R G SMITH,' J GARVEY,4 A SCHORK,2 B ENGLAND,3K D McCLATCHEY,3 AND J LINTON4

From the Department ofEnvironmental and Industrial Health, ' Department ofBiostatistics,2 and DepartmentofPathology,3 University ofMichigan, Ann Arbor, Michigan 48109, and Department ofBiology,4 SyracuseUniversity, Syracuse, New York 13210, USA

ABsTRAcr The relationship between metallothionein (MT), chronic exposure to cadmium (Cd),and renal function was investigated in 53 men who were occupationally exposed to Cd. The aimwas to determine if MT is a potential biological monitor for chronic exposure to Cd which wouldbe useful for preventing Cd nephropathy. In this study MT excretion, serum MT, and serumcreatinine concentrations were significantly higher in subjects with abnormal renal function whohad been exposed to Cd. MT excretion was also linearly related on an individual basis to proteinexcretion, f82-microglobulin (132-M) excretion, and cumulative time weighted exposure (dose).MT excretion was also a better predictor of dose than either ,82-M excretion or Cd excretion. Thefindings suggest that MT is a potential biological monitor for chronic Cd exposure that would beuseful for preventing Cd-induced nephropathy. Further studies of non-specific nephropathies andMT are needed to determine if MT is a specific indicator of proximal tubule function secondary tochronic exposure to Cd.

Metallothionein (MT) is a low molecular weight(<10 000) metal binding protein that was first iso-lated in 1957 by Margoshes and Vallee in equinekidney when they were searching for the mac-romolecule concerned in the accumulation of cad-mium (Cd) within the renal cortex.'MT has been identified in the cytosolic fraction of

liver, kidney, and other tissues in a wide variety ofanimal species and man. Synthesis of MT in thesetissues is inducible by cadmium, zinc, mercury, andcopper.2 The exact role of MT in metal metabolismis not known, but evidence suggests that it is con-cerned in cellular zinc homoeostasis3 and it may beresponsible for the nephrotoxicity of Cd observedafter chronic exposure.4

Chronic industrial exposure to Cd can result inearly renal dysfunction manifested as an increasedexcretion of 082-microglobulin and other low molecu-lar weight proteins (tubular proteinuria). Increasedexcretion of glucose, phosphate, and amino acidshas also been reported. A later effect on renal func-tion found in Cd workers is reduced glomerularfiltration and increased excretion of high molecularweight (>60 000) proteins.5

Received 30 July 1982Accepted 26 August 1982

305

Data from renal biopsies performed on peopleexcessively exposed to Cd suggest that nephropathydevelops when the concentration of Cd in the renalcortex exceeds 200 ,ug/g tissue.6 As indicated previ-ously, MT is concerned in the accumulation of Cd inthe renal cortex, but little is known about themechanism. One current proposal supported byresearch on experimental animals is that MT acts asa transport molecule for Cd from liver to kidney.7 Alow molecular weight protein such as Cd-MT wouldbe filterable by the glomerulus and fractionally reab-sorbed by the proximal tubule anatomically locatedin the cortex.8 If MT is concerned in thepathogenesis of renal dysfunction secondary to Cdexposure and is responsible for the accumulation ofCd in the renal cortex then it may be useful as abiological monitor for the prevention of Cd neph-ropathy in humans.To examine the relationship between MT and

occupational exposure to Cd, a study of 53 menoccupationally exposed to Cd from two separateplants (A and B) was conducted. A control group of56 men not exposed occupationally to Cd was alsostudied. The study was designed to examine the fol-lowing questions:

(1) Is there a relationship 3between cumulativetime weighted exposure (,ug/m years) to Cd, serum

on July 31, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Br J Ind M

ed: first published as 10.1136/oem.40.3.305 on 1 A

ugust 1983. Dow

nloaded from

http://oem.bmj.com/

Falck, Fine, Smith, Garvey, Schork, England, McClatchey, and Linton

MT, or urine MT?(2) Is there a relationship between renal function,

serum MT, and urine MT?(3) Is there a relationship between Cd in blood,

Cd in urine, MT in serum, and MT in urine?

Materials and methods

STUDY GROUPS

Plant APlant A produces refrigeration compressors withcopper fittings that are silver brazed. The brazingwire is 30-45% silver, 15-27% copper, 16-23%zinc, and 18-24% cadmium. A water based paste isalso used with the wire that contains potassium saltsof fluorine and boron. The refrigeration compres-sors are brazed either manually or by an automatedsilver brazing process. Both types of operation havelocal exhaust ventilation and, according to plantengineers, neither has changed since start up of therespective assembly line. The automatic line wasinstalled in 1968, before this all silver brazing wasdone on the manual line.9The air sampling data supplied by the Michigan

Department of Industrial Hygiene suggest thatexposures have been fairly constant over the yearson both operations. Twenty breathing zone samplescovering an 11 year period of operation of theautomatic line were used to derive an estimatedexposure of 39 + 7-8 (SE) ,ug/m3. Exposure on themanual line was founl to be significantly higher,110 + 25*5 (SE) ,ug/m ; 61 breathing zone samplesover a 21 year period were used for the calculationof this estimate. There was no statistically significantchange in exposure over the 21 year period on themanual line or the 11 year period on the automaticline; there was also no significant change of expos-ure with seasons (p > 0*10).9Employee work histories were obtained from the

personnel department from which a summary of thetime spent on the automatic versus the manual linewas calculated. The following formula was then usedto calculate cumulative time weighted exposure:

nTime weighted exposure = : Ei *T.

i = 1where

Ei = estimated inhalation exposure for a work area iTi = time worked in area i

Respiratory protection has never been required orused at this plant, so the previous estimates ofexposure were used without adjustment.

Thirty three male employees (group A) voluntar-ily agreed to participate in the study. This groupconsisted of 25 of 27 currently exposed workers and

eight of 13 previously exposed workers. Each par-ticipant was interviewed and administered a shortquestionnaire to obtain histories of hypertension,Fanconi syndrome, Goodpasture's syndrome, kid-ney infection, nephrolithiasis, analgesic and druguse, diabetes mellitus, diabetes insipidus, and smok-ing habits. From each participant a spot urine sam-ple was collected in a 125 ml metal freepolyethylene container. Each participant alsodonated 40 ml of blood by venepuncture in fourtrace element (certified Cd free) vacutainers withsuitable collection reagents for the specific analysis.9

Plant BPlant B produces Cd oxide and Cd sulphide fromflue dust collected by air pollution control equip-ment at non-ferrous smelters. The flue dust is ship-ped in bags by rail and each shipment is assayed forCd when received. The Cd containing material isroasted, dissolved in sulphuric acid, calcined, andpurified by precipitation and filtration. The solutioncontaining Cd is then pumped to the electrolytictank house where high purity Cd (95 %) is produced.From here the Cd is transferred to the foundrywhere it is melted and cast into bricks. Some of thebricks are reoxidised in gas heated retorts to makeCd oxide powder. Cd containing waste and residuefrom the retorts are used to produce yellow Cdsulfide pigment.10 - -

Exposure to Cd oxide dust occurs during thehandling of the Cd oxide powder and transport offlue dust between roasting, mixing, and calcining. Cdfume exposure occurs during the roasting, calcining,retort, and foundry operations. Exposure to Cd sul-phate mist occurs during the solution and electroly-tic tankhouse operations. Exposures to lead, arsenic,thallium, and indium also occur infrequently to afew individuals in isolated locations.'0

Smith et al'0 have studied this population exten-sively. Inhalation exposures have been estimatedusing area sampling data adjusted by a two step pro-cess that accounts for respirator use and the rela-tionship between area and breathing zone samples.The inhalation exposures reported by Smith et al'were used to estimate cumulative time weightedexposure.Work histories of each participant were estimate'd

using records of biological monitoring data thatlisted the department and job title at the time thesample was taken. Biological monitoring has beencarried out at the plant since about 1948. The inha-lation exposures and the work history were used tocalculate the cumulative time weighted exposure foreach participant.At the time the present study was conducted there

were 50 employees at the plant, and 20 men (group

306


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/

Metallothionein and occupational exposure to cadmium

B) voluntarily agreed to participate. A spot urinesample (40 ml) was collected from each participant.

Control groupForty one male controls not known to be occupa-

tionally exposed to cadmium were arbitrarilyselected from the University of Michigan for semi-quantitative and quantitative urine analysis. A 40 mlspot urine sample was collected from each. Fifteenadditional male controls were selected for meas-urement of serum /32-microglobulin (032-M).9

SAMPLE PREPARATION AND ANALYSISThe pH of each urine sample was checked after col-lection and adjusted above 5.5 where necessary inthose aliquots used for ,32-microglobulin analysis.Blood, serum, and urine samples used for semiquan-titative and quantitative analysis were stored at- 70°C. Serum and urine samples used for MTanalysis were stored at -20°C. All samples were

thawed at room temperature for subsequentanalysis.

Semiquantitative urine analysis for pH, protein,glucose, ketones, urobilinogen, bilirubin, and bloodin urine was done using Chemstrip 7 (Biodynamics,Indianapolis, USA). Quantitative urine analysis wasdone for the following: Cd, total protein, albumin,/32-M, glucose, creatinine, MT, specific gravity, andosmolality. Cd was analysed on a Varian 475 (PaloAlto, CA) atomic absorption spectrophotometerusing NIOSH method physical and chemical analyti-cal method 173.11 Total protein was analysed usingthe Quant test (Quantimetrix, Hawthorne, CA)based on the method of Bradford.'2 Albumin was

measured using the Gelman (Laboratory ProductsDivision, Ann Arbor, MI) protein electrophoresissystem. 132-M was determined by radioimmunoassay

using the Phadebas 82-micro test (Pharmacia Diag-nostics, Uppsala, Sweden). Glucose and creatininemeasurements were performed on a Beckman Astra8 (Beckman Instruments, Brea, CA). Osmolalitywas done using a Wescor 5100 B vapour pressureosmometer (Wescor, Inc, Logan, UT), and specificgravity was measured using a TS meter (catalogueNo 10400, American Optical, Keene, NH). MT was

measured by a radioimmunoassay methoddeveloped by Vander Mallie and Garvey.'3 14

Serum samples were analysed for creatinine,f32-M, and MT by the respective instruments andmethods outlined above. Blood Cd was measuredusing a Varian 475 atomic absorption spec-trophotometer equipped with a graphite furnace.

DATA ANALYSISThe results of the quantitative urine analysis, includ-ing glucose, total protein, f32-M, MT, and Cd were

transformed for comparative purposes by beingexpressed per gram of creatinine. This transforma-tion did not alter any of the statistical distributionsor any of the statistical conclusions for thehypotheses investigated. A logarithmic transforma-tion was then performed on the creatinine adjustedvariables since they approximated a log-normal dis-tribution. Linear regression analysis was used todetermine the statistical relationship between MT(the dependent variable), cumulative time weightedexposure (,ug/m years), total protein, f32-M, glu-

cose, creatinine, and Cd in groups A and B sepa-

rately, since these groups differed3 significantly(p < 0-001) in Cd excretion andug/m years (table1). This was predictable since group B had highercumulative time weighted exposures than group A(table 2), and different methods for estimatingexposure years were used for each group. The two

Table 1 Characteristics ofcontrol group and groups A and B

Variable Controls Group A p value Group B p value*(n = 41) (n = 33) (n = 20)

Aget 40 + 13 50 + 6 <0-001 50 ± 12 0-001Height, cmt 178 ± 7 175 + 8 0-26 175 ± 9 012Weight, kgt 75 + 10 87 + 13 <0-001 87 + 20 0-004UrinepHt 5-8 ± 0-9 5-0 + 0-9 <0-001 6-4 + 0-9 0-03Specific gravityt 1-021 + 0-01 1-020 ± 0-004 0 37 1-014 + 0-01 0-01Osmolalityt 843 + 292 724 + 163 0-04 599 + 225 0-002Creatininet g/100 ml 0-17 ± 0-08 0-14 + 0-05 0-06 0-11 + 0-05 0-004Cadmiumt 4ggg creatinine 12-4 + 1-6 23-7 + 1-8 <0-001§Metallothioneint

,gticreatinine 2-4 + 2-2 40-8 ± 3 33 + 3 0-44§mg/M3 years* 593 + 2-5 1994 + 2-6 <0-001§

Student t test (groups A and B versus control group).tMean + SD.tGeometric mean ± SDa (SDa is antilog of SD in log scale).§Group A versus group B.

307


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/


Table 2 Quandtative urine analysis results expressed per gram creatinine and cumulatve tme weighted exposure data;cadmium exposed groups A and B*

Subjectst mg Glucoseig mg Proteinlg mg Albununig jpg ,f8-Mlg pg MTlg pg Cd/g gm/r3 yearscreatinine creatinine creadinne creahnine creatinine creatinine

12t34t56789101112t131415t1617181920t2122232425t262728293031323334t3536t3738394041424344454647484950515253t

495252774354504191384324728901102643365741405534618659454027

723326026

2175349555732485241192504062473165

1352633125

3397

3347811114843324121103274351728

17011026573628

355201568534

32130131521

660163310431711421369633621247353

23041372138

1461753

382095

261

114

266

91

197

159

74644

3496227879322521631657

6103456

8920713401798628

1849779

440428109

5366730131127

2824116

653117

195851207713018111263672

397507150861447067

10431142000

6325

2270

4487332462178940543327657398353

41869

91377832402684011161774563414666396516255729103310625S

235

282332561514

116

1295810131110105

171813121213111710241021171061207914189

11212619414117241615272775122123122466101348

1164904955201102010861079428351217486190146811251591

* 18773708162

132044735610657001996230478722725162942257444418962738277143033342263429420751938968530350

2543174970113326

2000773016402747

*Reference tolerance limits: 130 mg glucose/g creatinine, 172 mg protein/g creatinine, 629 A& ,82-M/g creatinine.Note: Control geometric mean ± SEa (SE' is antilog of SE in log scale): 47 mg glucose/g creatimne ± 1- 1, 39 mg protein/g creatinine ± 1-0,83 p; ,MNVg creatinine ± 1*2, 2*4 pg MT/g creatimne ± 1*2.tSub ects 1-33 group A, subjects 34-53 group B.tSubjects classified as having abnonnal renal function-that is, exceeded reference tolerance limits for two of three urine variables or werealbuminuric and exceeded one upper reference tolerance limit.

groups did not differ in MT excretion (p = 0-44).For each regression equation the T-statistic was cal-culated and the level of significance determined.To examine the relationship between MT and

renal status in the study group, tolerance limits(0.95) were calculated from the control group datafor each variable. If a subject exceeded the upper

limit on two or more of the following urinevariables-glucose, total protein, and f32-M-or hadalbuminuria and exceeded the upper limit on one ofthe urine variables his renal function was classifiedas abnormal. This decision rule is supported by evi-dence that Cd can effect both proximal tubule andglomerular function.5 The Student t test was used to

308


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/


Table 3 Renal functon and signifiance ofbiological variables, cadmium exposed groups A and B

Group A Group B Groups A and B

Abnormal Normal p-value Abnormal Normal p-value Abnormal Normal p-value(n = 6) (n = 26) (n = 3) (n = 16) (n 9) (n =42)

Age* 54 + 1 49 + 1 0-08 57 ± 2 48 3 0-22Urine metallothioneint 69 + 0-4 36 + 1-2 0-09 79 + 1-2 25 ± 1-3 0-07 72 + 1-3 31 ± 1-2 0-01

Qtg/g creatinine)Unne cadmiumt 17 ± 1-4 11 ± 1 0.09 37 + 1-2 20 + 1.1 0-07 - - -

(jg/g creatinine)Unne proteint 275 ± 1-3 39 ± 1-2 <0-001 186 ± 1-4 49 ± 1-2 0-007

(mg/g creatinine)Uine c,r-Mt 6310 ± 2-4 79 ± 1-4 <0-001 3236 ± 2-6 162 ± 1-4 0-005

(ytg/g creatinine)Serum metallothionein* 3 ± 0-62 2 ± 0-16 0-02 - - - - - -

(ng/ml)Serum creatinine* 1-4 ± 0-11 1-06 ± 0.04 0.001 -(mg/100 ml)

Serm B -microglobulin* 2-3 + 026 2-1 0-17 0-58 -oizg/ml6

tGeometric mean -+SEa (SEa antilog of SE in the log scale).*Arithmetic mean -SE.

test the significance of the difference in MT excre-

tion in subjects with abnormal versus normal renalfunction.

Linear regression analysis and analysis of varianceprocedures were used to examine the effect, if any,

that smoking habits had on Cd excretion, and thatage had on protein, MT, and ,82-M excretion.

Results

Table 1 presents the general characteristics of thecontrol group and groups A and B. There is a

significant difference in the average age and weightof groups A and B versus controls. A significantdifference also exists between group B and the con-trol group in average specific gravity, osmolality,and urinary creatinine concentration. Group A andthe control group differ significantly in mean osmol-ality. No significant difference (p > 0-40) was foundwhen the mean excretion of glucose, protein, and,82-M in the control group (bottom table 2) wascompared with the mean excretion in subjects withnormal renal function in groups A and B (table 3).The quantitative urine analysis results for glucose,

total protein, ,82-M, MT, and Cd expressed per gramof creatinine are listed by subject in table 2. Thereference tolerance limits calculated from the con-trol group data are located at the bottom of table 2.The reported reference tolerance limits agree withwhat would be expected for unexposed normal sub-jects, assuming 1-2 g of creatinine is excreted eachday.'s Blood samples collected from group A wereanalysed for creatinine, f32-M, MT, and Cd; resultsby subject are reported in table 4. Reference toler-ance limits determined from control serum samplesare located at the bottom of table 3.Cd excretion (log ,uglg creatinine) differed

(p = 0.06) between smokers, non-smokers, and

ex-smokers. Smoking habits, however, did not haveany significant effect on the relationship between Cdexcretion and the other variables under study when

Table 4 Quantitative analysis ofserum, cadmium exposedgroup A *t

Subject Serum creannine Serum Serum(mgIl00 ml) microglobuln metalothionein

(11/gi i) (nglml)1 1-0 1.3 1-42t 1-6 1-7 4-93 1-3 2-2 1-64 1.1 1.9 1-65 1-0 1-1 <1-06 0.9 2.0 2-47 1-1 3-3 1-88 1-2 2-8 1-59 1.1 1-5 1-610 0.9 2-0 1-611 1-0 1-9 1-012t 1-8 2-4 1-613 0.9 1-2 1-714 1.0 3-4 1-615* 1-5 3-0 1-716 1-3 2-2 1-217 1.1 2-6 1-818 1-6 1.6 2.819 1-1 1-5 2-320 1-3 1-8 3.721 0-8 3-0 1-322 1-0 2-2 1-323 1.1 3.0 2*424 1i0 0.2 1-825 1-2 3-2 4-326 1.1 1-6 2-127 1.1 2-0 1*628 0-9 1-3 1-629 1-1 1-6 4-230 1.3 3.4 1-831 0.5 1-9 3-532 1-5 4-1 3.533 0.9 3-0 1-7

Reference tolerance limits: creatinine 1-3 mg/100 ml, #,8-M 3-9

tAIl subjects blood Cd < 5 ng/ml.*Subjects with mildly raised serum creatinine concentrations whoare also classified as having abnormal renal function according todecision rule at bottom of table 2.

309


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/


it was included as an independent factor in multifac-tor regression analysis. Age had no significant effecton any of the statistical relationships when includedas an independent factor.

After comparing the values in table 2 with therespective reference tolerance values, it becomesapparent that subjects 2, 4, 12, 15, 20, 25, 34,'36,and 53 have undergone a change in renal functionconsistent with Cd exposure. Abnormal results arealso reported for subjects 30 and 50. Subject 30,however, has a history of kidney infection and sub-ject 50 has diabetes mellitus. These pre-existingmedical conditions can lead to renal dysfunction, sothese two subjects were not used in the statisticalanalysis of the study data. Subjects 2, 4, 12, 15, 20,and 34 are albuminuric, and in subjects 2, 12, and15 the serum creatinine concentration is raised(table 4). Subject 15 also has a history of neph-rolithiasis. Several investigators have reported anassociation between nephrolithiasis and exposure toCd.5 There is no significant difference in the averageage of subjects with abnormal versus normal renalfunction within groups A or B (table 4).The relationship between MT excretion and renal

status was examined by comparing subjects withabnormal versus normal renal function and MTexcretion within each group and after pooling bothgroups. Pooling of groups A and B was done withsome caution since the groups differed with respectto Cd excretion and cumulative time weightedexposure; they did not, however, differ significantlyin MT excretion. MT excretion in groups A and Bcombined is on the average significantly higher(p = 0-01) in subjects with abnormal versus normalrenal function (see table 3). MT excretion is alsohigher within each group for subjects with abnormalversus normal renal function. The level ofsignificance (group A, p = 0-09; group B,p = 0-07), however, is much less, and this is possiblydue to the smaller sample size within each group.Serum creatinine (p = 0.001) and serum MT(p = 0.02) are, on the average, significantly higherin the subjects with abnormal versus normal renalfunction within group A (table 4). MT excretion islinearly related to protein and 32-M excretion ingroups A (p = 0-001, p = 0.004; figs 1, 2 respec-tively) and B (p = 0-008, p < 0*001; figs 1, 2respectively). These relationships suggest that MTconcentrations in urine or serum are related to renalfunction.Excretion of MT is also linearly related to cumula-

tive time weighted exposure in group A (p < 0*001,fig 3) and group B (p = 0-09, fig 3). Cd excretion isnot significantly linearly related to cumulative time

2weighted exposure in group A (r = 0*06, p = 0.20)or group B (r = 0-04, p = 0.44). Excretion of MT

and Cd are linearly related both in group A(p = 0-009, fig 4) and group B (p = 0.07, fig 4). Itshould be indicated that in fig 4 the lowest point forgroup A shifts the line somewhat.2If this point isremoved the correlation improves (r = 0-32); there

2 s-S

225 -

2 00-

0)c

_ 1 75-0

c 1 50-dic0-C0=1 25-

E

1 00-0

0 75-

0 50-

f,

/Group B (*&)

,// Group A l-)

0 j y = 076 + 047x±0 333r2=0331|

a - --- 0.22+071x± 0 I37A r 2= 0.35

1 00 1 50 200 2-50 3 00log mg protein /g creatinine

Fig 1 Urinary excretion ofmetallothionein and protein,cadmium exposed group A and group B.

2 50-

2 25-

_@ 200-

e 1 75 -

& 1 50-.o

o 125-2-aE 100-

° 075

050-

Group B (A)

* -, Group A (-)

-20* 25y=051+039x±031

5~~ r2= 0o50

15O 2 00 250 300 350 4 00 4 50 5 00iog ug 82 -m(croglobulin/g creatinine

Fig 2 Urinary excretion ofmetallothionein and/32-microglobulin, cadmium exposed group A and group B.

310

A

11

0

0I A

2 000000

a


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/


2 50-

2-25-

,C 2 00-

"175-a,c

c 1 50-

E0E 100-

3'w

G 75-

0 50-

Group A (s)

Group B (l-

..V2 7,,' +~H~

/ *,o2 * ~~~Y= -0 *11 +0 62 X! 0 30

I r2=0o40I/,0=y- 20 0-52X 0 41|-=

1 50 200 250 3C0 350 400 450

tog g yeorm3

Fig 3 Urinary excretion ofmetallothionein and cumulativetme weighted exposure (uglm3 year), cadmium exposedgroup A and group B.

2 50-

2-25-o

,1 75-

c

g150pu

° 125-E

c1:; 100-

/Group A (e)

,-Group B *

y=-068 0-85x ±0 35r2= O 21

y=036 t 0 82Xx! 0 41r2 =018

0 75 1 00 1 25 1 50 1 75 2 00log ug .codmium/g creotinine

Fig 4 Urinary excreton ofmetallothionein and cadmium,group A and group B.

is no valid reason to remove it, however. These rela-tionships suggest that MT excretion is more highlycorrelated with cumulative time weighted exposurethan Cd excretion and that MT excretion and Cdexcretion are related.Serum MT in group2 A was not significantly

related to serum ,82-M (r = 0 01, p = 0-60), serumcreatinine (r = 0O07, p = 0-15), or cumulative timeweighted exposure (r = 0-08, p = 0 11). MT excre-tion was not significantly linearly related to serumMT (r' = 0-07, p = 0-13). Blood Cd concentrationswere less than the lower detection limit of 5 ng/mlfor all subjects of group A, thus the relation, if any,between serum MT and blood Cd could not be

examined. To determine if MT excretion is a betterindicator of exposure than ,82-M excretion, in groupsA and B partial correlations were computed.', Usingthis technique the following correlations werefound: for MT excretion and cumulative time weigh-ted exposure for groups A and B r = 0-29 and 0-19when ,82-M excretion was partialled out, and for832-M excretion and cumulative time weightedexposure for groups A and B, r2 = 0-02 and 005when MT was partialled out. These correlationssuggest that MT excretion is a better predictor ofcumulative time weighted exposure.A trend was observed between mean levels of MT

excretion and proportion of subjects with an abnor-mal renal status in groups A and B combined. Logis-tic regression analysis was done on the proportion ofsubjects with an abnormal renal status and MTexcretion, the level of significance is 0*02 and the rvalue 096. This finding however, should be inter-preted cautiously, since no subjects with abnormalrenal function were found in the second stratumwhen MT excretion in groups A and B were categor-ised into five strata and the mean MT excretion wasused as the independent variable in each strata.

Discussion

In group A 18% and in group B 15% of the subjectswere found to have a change in renal function con-sistent with exposure to Cd. The lower prevalence ofrenal dysfunction in group B is not compatible withthe higher exposures in this group. This could poss-ibly result from the removal of subjects with abnor-mal renal function, selection bias, or the mis-classification of exposure and work histories. Elliset al studied this entire group and found that theprevalence of renal dysfunction was 23% among thecurrently exposed workers (n = 51).'7

Age, analgesic abuse, and specific medical condi-tions (see methods) were possible confounding fac-tors in the study, since they can have an effect onrenal function. Age, in this study, did not have asignificant effect on any of the linear relationshipsbetween the variables. Subjects with abnormal renalfunction and identified (questionnaire) medical con-ditions that could affect renal function were not usedin the statistical analysis of the study data. No casesof analgesic abuse were identified by questionnaire.Smoking habits were also a potential confoundingfactor, since Cd excretion is significantly higher insmokers.5 There was a significant difference in Cdexcretion between smokers, non-smokers, andex-smokers; however, smoking habits had nosignificant effect on any of the linear relationshipsregarding Cd excretion and the other variables.MT excretion was significantly higher, on the

a'I(V

311

0-75

0, 50..,,


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/


average, in subjects with an abnormal versus normalrenal status in groups A and B and after poolinggroups A and B (table 3). Age did not differsignificantly in subjects with abnormal versus nor-mal renal function in groups A and B, suggestingthat the increased excretion of MT is not an ageeffect but rather is related, on a group basis, to renalfunction. MT excretion was also linearly related toprotein excretion (fig 1) and,(2-M excretion (fig 2)in groups A and B. This finding seems significant,given that protein excretion and 132-M excretion areindicative of renal function in workers exposed toCd.'8 Tohyama et al also found that on a group basisMT excretion was significantly higher in those withrenal tubular dysfunction secondary to excessiveenvironmental Cd exposure.19Serum MT and serum creatinine concentrations

were also significantly higher in subjects withabnormal versus normal renal function in group A(table 3). This relationship suggests that theincreased serum MT concentrations are related to adecrease in renal function, since serum creatinineconcentration is a reliable predictor of renal func-tion.20 These findings also suggest that monitoringMT in the urine or serum may be useful for prevent-ing renal dysfunction resulting from excessiveexposure to Cd. The presence of MT in the serum ofCd exposed workers is a significant finding. Theserum MT concentration could result from leakageof MT back into the plasma secondary to degenera-tion or necrosis of renal parenchymal tissue, or both,or MT may be acting as a transport molecule for Cdfrom the liver to the kidney. If the second hypothesisis true those individuals having increased concentra-tions of serum MT would accumulate Cd in the renalcortex more rapidly and, therefore, be at greater riskof developing Cd nephropathy.The linear relation between MT excretion and

cumulative time weighted exposure (group A,r2= 040; group B, r2 = 0.17) implies that MTexcretion is dose dependent. The variability thatexists, especially within group B, is to be expectedand is the combination of a small sample size, esti-mation of cumulative time weighted exposure, andinterindividual differences in absorption, distribu-tion, metabolism, and excretion.2' Group A wasfound to be more homogeneous than group B whencomparing the type of exposure, exposure levels,and characteristics, which also contributed to thedifferences in variability between the groups and therelationships that were investigated. Cd excretionwas not dose dependent (cumulative time weightedexposure) in either group A or group B. This findingis consistent with studies that have reported thaturinary Cd is not related to chronic exposure orbody burden, but rather becomes raised when renal

dysfunction develops.5 ,32-M excretion was also notsignificantly correlated, on an individual basis, withcumulative time weighted exposure. Thus on anindividual basis the findings in the study suggest thatMT is a better indicator of cumulative time weightedexposure than Cd or ,32-M and, therefore, a betterpredictor of body burden.The linear relation between MT and Cd excretion

agrees with the findings of other investigators, andsuggests that MT is concerned in the renal handlingof Cd. Chang et al and Tohyama and Shaikh alsofound that urinary MT and urinary Cd in exposedhumans were significantly related.2223Serum MT concentration was not significantly

linearly related to cumulative time weighted expos-ure or MT excretion, which is also in agreement withthe findings of Chang et al.22 There are several poss-ible explanations for this; the most feasible beingthat a small molecular weight protein such as MTwould be rapidly cleared by the kidney, and only incases where renal function is impaired would theconcentrations be significantly raised.The findings in this study suggest that MT is

related to the nephrotoxicity of Cd. MT was linearlyrelated to protein excretion, ,82-M excretion, andcumulative time weighted exposure. The concentra-tions of MT were also significantly higher in subjectswith abnormal renal function, consistent with Cdexposure. This suggests that MT is a potentialbiological monitor for Cd exposure which would beuseful for preventing Cd nephropathy in man.Further studies need to be done to define the kine-tics of the renal handling of MT and also to deter-mine if MT is a specific indicator of proximal tubuledynsfunction secondary to Cd exposure. Before MTcan be used as a biological monitor for Cd exposure,studies of non-specific nephropathies need to beunderaken to validate that MT is not a non-specificindicator of proximal tubule dysfunction such as132-M.

We thank Ms Karen Wilkerson, Ms Barb Brownlee,and Mr John Jarrett for their technical help and MsJean Tashian, Ms Pam Chlebek, Ms DoreenMcCann, and Mrs Maryann Bymes for assisting withthe preparation of the manuscript. We would alsolike to thank Dr Arthur Vander for his support ofthis study.

References

'Kojima Y, Kagi JHR. Metailothionein. Trends In BiochemicalSciences 1978;3:90-2.

'Webb M. The chemistry, biochemistry and biology ofcadmium.Topics in environmental health. Vol 2. New York: Elsevier/North-Holland, 1979;205-20.

312


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/

Metallothionein and occupational e-xposure to cadmium

3Menard MP, McCormick CC, Cousins RJ. Regulation of methal-lothionein biosynthesis in intestine and liver by dietary zinc.Fed Pro March 1980;39:880.

4Cherian GM, Goyer RA, Richardson-Delaquerriere L.Cadmium-methallothionein induced nephropathy. ToxicolAppl Pharmacol 1976;38:399-408.

'Friberg L, Piscator M, Nordberg GF, Kjellstrom T. Cadmium inthe environment. 2nd ed. Cleveland: Chemical Rubber Com-pany, 1974:102-5.

World Health Organisation task group. Health criteria for cad-mium. Geneva; WHO 1977. (Summary.)

7Cherian MG, Vostal JJ. On the role of cadmium binding proteinin the transport and excretion of cadmium in the rat. ToxicolAppl Pharmacol 1974;29:141.

Vostal JJ, Cherian MG. Effects of cadmium methalothionein onthe renal tubular transport of sodium. Fed Proc March1974;33:1737.

Falk FY, Fine UJ, Smith RG, et al. Occupational cadmiumexposure and renal status. Am J Ind Med 1983;4:541-9.

10 Smith TJ, Anderson RJ, Reading JC. Chronic cadmium expos-ures associated with kidney function effects. Am J Ind Med1980;1:319-37.

US Department of Health, Education and Welfare. NationalInstitute for Occupational Safety and Health. NIOSH manualofanalytical methods. Washington: DHEW 1974. (HEW pub-lication No (NIOSH) 75-121, 173:1-8.)

12 Bradford MM. A rapid and sensitive method for the quantitationof microgram quantities of protein utlizing the principle ofprotein-dye binding. Anal Biochem 1976;72:284-54.

13 Vander Mallie RJ, Garvey JS. Production and study of antibody

produced against rat cadmium thionein. Immunochemistry1978;15:857-68.

"Vander Mallie RJ, Garvey JS. Radioimmunoassay of metal-lothioneins. J Biol Chem 1979;254:8416-21.

5 Henry JB, ed. In: Todd, Sanford, Davidsohn. Clinical diagnosisand management by laboratory methods. 16th ed. Philadel-phia: WB Saunders, 1979: 2094-7.

16 Neter J, Wasserman W. Applied linear statistical models.Homewood: Irwin, 1974:393-415.

7 Ellis KJ, Morgan WD, Zanzi I, Yasumura S, Vartsky D, CohnSH. In vivo measurement of critical level of kidney cadmium:dose effect studies in cadmium smelter workers.Am J Ind Med1980;1:339-48.

"Kjeilstrom T, Evrin P, Rahnster B. Dose-response analysis ofcadmium-induced tubular proteinuria. Environ Res1977;13:303-17.

9Tohyama C, Shaikh ZA, Nogawa K, Kobayashi E, Honda R.Elevated urinary excretion of metallothionein due to environ-mental cadmium exposure. In: Toxicology. Vol 20. New York:Elsevier/North Holland, 1981:289-97.

Hook JB. Toxicology of the kidney. Target organ toxicologyseries. New York: Raven, 1981:1-29.

21 Gompertz D. Assessment of risk by biological monitoring. Br JInd Med 1981;38:198-201.

22 Chang CC, Vander Mallie RJ, Garvey JS. A radiommunoassayfor human metallothionein. Toxicol Appl Pharmacol1980;55:94-102.

23Tohyama C, Shaikh ZA. Urinary metallothionein as an index ofcadmium-induced renal dysfunction in humans. Fed Proc April1982;41:1226.

313


http://oem.bm

j.com/

Br J Ind M


ugust 1983. Dow

nloaded from

http://oem.bmj.com/

metallothionein and occupational to cadmium · metallothionein (mt) is a low molecular weight (

Documents