High-pressure liquid chromatography for isolation of clastogenic agents from urine

Mutation Research, 147 (1985) 171-177 171 Elsevier

MTR 08577

High-pressure liquid chromatography for isolation of clastogenic agents from urine

J. Randall Curtis and Bruce P. Dunn Environmental Carcinogenesis Unit, British Columbia Cancer Research Centre, 601 West l Oth Avenue, Vancouver, B.C. V5Z 1L3

(Canada)

(Received 25 September 1984) (Revision received 10 January 1985)

(Accepted 14 February 1985)

Summary

Small columns of XAD-2 resin have been widely used to extract and concentrate mutagenic materials from urine. Using analytical HPLC and assays for clastogenicity with Chinese hamster ovary cells, we found that small columns of XAD-2 resin (1.5 ml bed volume) retain only a small percentage of organic material and undetectable amounts of genotoxic activity in urine samples. Increasing the size of the XAD resin bed resulted in better recoveries, but much organic material was still lost by overloading of the column. In contrast, when urine was acidified and chromatographed by preparative reversed-phase HPLC using large-bed-volume (500 ml) commercial columns, retention of hydrophobic organic material from urine was excellent. Subsequent stepwise elution of the column with increasing concentrations of acetone produce 3 fractions of organic material of increasing hydrophobicity. When urine from smokers was analysed, all 3 fractions contained material which was clastogenic to Chinese hamster ovary cells. The procedures developed are suggested as a new general purpose approach to the isolation of genotoxic materials from urine.

Mutagenic activity has been identified in the urine of several groups thought to be exposed to carcinogens. These groups include cigarette smokers (Yamasaki and Ames, 1977; Falck et al., 1980; Putzrath et al., 1981; Aeschbacher and Chappuis, 1981; Laires et al., 1982), passive cigarette smokers (Bos et al., 1983), pharmacy personnel handling mutagenic antineoplastic drugs (Minnich et al., 1976; Nguyen et al., 1982; Ander- son et al., 1982; Bos et al., 1982), and patients undergoing different chemotherapy regimes (Lega- tor et al., 1975; Guerrero et al., 1979).

Since urine is generally too dilute to be used directly in short-term genotoxicity assays, a concentration procedure is required. When mutagenicity assays based on histidine auxotrophs of bacteria

are used, purification of urine to remove histidine is also required to avoid artifacts (Gibson et al., 1983). Perhaps the most common and successful procedure used to concentrate organic material from urine for mutagenicity testing is the use of small (1.5 ml bed volume) columns of XAD-2 resin, as described by Yamasaki and Ames (1977). Although widely applied, this procedure can suffer from poor recoveries of mutagenic material from urine. Yamasaki and Ames (1977) note that while the amount of mutagenic material retained by a column increases as the amount of urine passed through the column is increased, the recovery of mutagenicity per unit volume of urine when 500 ml of urine is passed through a column is less than 15% of the recovery when only 10 ml of urine is

0165-1161/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)

172

used. Beck et al. (1982), using sister-chromatid exchange as an assay system, found that urine which had already been passed through a small (1.5 ml bed volume) XAD-2 column, yielded simi- lar amounts of genotoxic activity when it was passed through a second and then a third column.

In an attempt to develop a more selective and sensitive screening procedure, we used preparative reversed-phase high-pressure liquid chromatography (reversed-phase HPLC) to concentrate and fractionate urine samples. We applied the urine concentrates to cultured Chinese hamster ovary (CHO) cells and then analysed cells for chromosome aberrations. We compared the reversed- phase HPLC procedure to the XAD-2 resin concentration procedure in terms of the amount of organic material retained from a given sample (evaluated by analytical HPLC), and in terms of the level of chromosome aberrations induced by the retained material.

Materials and methods

Creatinine assay The creatinine concentration of each urine sam-

ple was determined using a modification of the assay described by Iosefsohn (1982). To 1.5 ml 10 mM picric acid was added 0.5 ml 600 mM NaOH and 10 /~1 urine, water, or creatinine standard. Samples and standards were measured against blanks containing the 10/~1 water at 520 nm in a double-beam spectrophotometer.

Urine concentration by XAD-2 resin columns Two XAD-2 column procedures were used: (1) The procedure described by Yamasaki and

Ames (1977). Briefly, 500 ml urine was passed at a flow rate of 1-2 ml /min through 0.7 g (dry weight) XAD-2 resin (bed volume 1.5 ml) contained in a 0.7 c m x 10 cm glass column. Urine wetting the packing was removed by briefly blowing nitrogen through the column without allowing the packing to dry. Unretained material was washed from the column with 1.5 ml water, and water was removed from the column using nitrogen. Organic material was eluted from the column with 10 ml acetone.

(2) A scaled-up procedure, carried out in the same manner as procedure (1) and using the same volume of urine, except that a 4.5 cm X 60 cm

column with 25 g (dry weight) resin was employed. The column was washed with 55 ml water, then eluted with 350 ml acetone.

Acetone eluates were evaporated to dryness on a rotary evaporator with a water-bath temperature of 40°C. The residue was dissolved in 10 ml of 20% DMSO and stored at - 2 0 ° C .

Urine concentration by preparative high-pressure liquid chromatography

Except where noted, urine samples were acidified to pH 3.0 with phosphoric acid, then filtered through coarse fluted filter paper to remove a gelatinous precipitate that commonly formed on acidification. Reversed-phase preparative HPLC was carried out at a flow rate of 100 ml /min with helium-degassed solvents on a Waters 500A Prep LC system (Water Scientific, Milford, MA) equipped with a Waters Prep 500A/C18 reversed-phase column (void volume approximately 500 ml). Each day prior to use, the column was cleaned by eluting it with 1 1 acetone followed by 2 1 2.5 mM phosphoric acid.

Urine samples ranging from 350 to 700 ml were loaded onto the column by placing the HPLC pump solvent inlet line in the urine sample and using the HPLC pump to pass the urine through the column. The urine was then followed with 1 1 of 2.5 mM phosphoric acid which was discarded. Organic material retained on the column was then eluted in 3 fractions using (1) 900 ml of 5% acetone/95% 2.5 mM phosphoric acid, (2) 500 ml of 15% acetone/85% 2.5 mM phosphoric acid, and (3) 500 ml of 50% acetone/50% 2.5 mM phosphoric acid. The effluent of the column was moni- tored with a refractive index detector, and fraction cut points were established when the detector baseline first started to shift in response to increased concentrations of acetone for each eluant. Fraction 1, resulting from the pumping of 900 ml of 5% acetone, comprised only approximately 500 ml, as the acetone content of the first 400 ml of eluant is adsorbed by the reversed-phase column. This effect causes the refractive index change indi- cative of the first arrival of acetone in the column effluent to be 400 ml later than would be expected if the column did not adsorb acetone. Fractions 2 and 3 do not suffer from this effect since the column is fully saturated with acetone during their

elution. These fractions were therefore 500 ml each.

The fractions were rotary evaporated to remove acetone and then neutralized and freeze-dried. Freeze-dried residues were dissolved in 10 ml water, containing in some cases up to 25% DMSO, and then stored at - 2 0 ° C . It has since been determined that the use of DMSO is rarely necessary in order to redissolve the organic material.

Analytical high-pressure liquid chromatography Urine fractions were characterized by analytical

HPLC using a Whatman PSX 5/25 ODS-3 reversed-phase column, and an eluting solvent of 5% methanol/95% 0.1% phosphoric acid for 5 min followed by a gradient rising by 5% methanol/min. A flow rate of 1 ml /min was employed, and detection was by UV absorption at 220 nm. In order to allow easy comparison of the relative amounts of organic material in each fraction, samples were individually diluted to a creatinine equivalence concentration of 2 mg /ml (i.e., 1 ml of diluted sample contained the organic material originating from a volume of urine containing 2 mg creatinine); 5-/~1 aliquots were then chromatographed.

Cell cultures and assay for chromosome aberrations CHO cells were grown as previously described

by Stich et al. (1981). For experiments, 40-60% confluent cultures were grown on 22-mm square coverslips in 35-mm tissue culture dishes. Urine fractions were diluted to a concentration of 4.0, 2.0, 1.0, 0.5 and 0.25 mg/ml creatinine equivalence in Eagle's minimal essential medium (MEM) supplemented with 2.5% fetal calf serum. In the terminology used in this paper, 4.0 mg/ml creatinine equivalence means that per millilitre final volume, the extracts applied to the cells contained the organic material isolated from a volume of urine originally containing 4 mg creatinine. CHO cells were exposed to urine fractions for 3 h at 37°C, after which the urine concentrate was removed and the cells washed with MEM and cultured a further 16 h in MEM supplemented with 10% fetal calf serum. Chromosome aberrations were scored after a 3.5-h colchicine treat- ment, as described previously (Stich et al., 1981). For each coverslip a minimum of 100 metaphase

173

plates were analysed. A cell was scored as positive if it contained at least one chromatid or chromosome break or exchange.

Results

Comparison of the material in urine retained by XAD-2 resin and preparative reversed-phase HPLC

3 1 of urine were collected from a heavy smoker over a 3-day period (excluding morning void) and were pooled. 500-ml aliquots of this pooled sample were subjected to different types of preparative chromatography. Fig. 1 shows the analytical HPLC profile of the original urine compared with the material isolated by the small-scale and large-scale XAD-2 column chromatography. The amounts of material chromatographed have been adjusted so that relative peak heights of components in the original urine and in the extracts are a direct measure of recovery. It is evident that chromatog-

T r 10 20 30 Minutes

Fig. 1. Extraction of organic material from urine by chromatography on XAD-2 resin. Analytical HPLC was performed as described in Materials and Methods. The amount of material injected corresponds in each case to that recovered from a volume of urine containing 10/zg creatinine. (a) Urine prior to chromatography; (b) extract from small (0.7 g dry weight resin) XAD-2 column; (c) extract from large (25 g dry weight resin) XAD-2 column.

174

raphy on the small XAD-2 column, as described

by Yamasaki and Ames (1977), is inadequate in

isolating more than a small fraction of the organic

material in a 500-ml urine sample. Recoveries from

the large-scale XAD-2 co lumn were substant ia l ly

better, but large amounts of material originally present in the ur ine were not represented in the

organic concentrate. Analyt ical HPLC of the urine

and water washes passed through the co lumn re-

vealed that the losses of organic material were due

to overloading of the co lumn rather than to an

- ~

,, t I

--6- " I

r r 0 I0 20 30

Minutes

Fig. 2. Extraction of organic material from urine by preparative reversed-phase chromatography at neutral pH. Analytical HPLC was performed as described in Materials and Methods. The amount of material injected corresponds in each case to that recovered from a volume of urine containing 10 /.tg creatinine. DMSO peak in (c) and (d) comes from solution used to dissolve extracts after chromatography and freeze-drying. (a) Material washed from the column with water prior to acetone elution; (b) material eluted from the column with 5% acetone; (c) material eluted from the column with 15% acetone; (d) material eluted from the column with 50% acetone.

inabi l i ty to subsequently elute organic material from the co lumn with acetone (data not shown).

Fig. 2 shows the results of analytical HPLC

analysis of ur ine fractions prepared using prepara-

tive HPLC without acidification of urine or elu- t ion solvents. Fig. 2a shows the material eluted

with water from the co lumn prior to acetone elu-

t ion (a fraction which is normal ly discarded and

not subjected to genotoxicity testing). Fig. 2b, c

all i

_j

so

r 0 10

Minutes 20 30

Fig. 3. Extraction of organic material from urine by preparative reversed-phase chromatography at acid pH. Analytical HPLC was performed as described in Materials and Methods. The amount of material injected corresponds in each case to that recovered from a volume of urine containing 10 #g creatinine. DMSO peak in (c) and (d) comes from solution used to dissolve extracts after chromatography and freeze-drying. (a) Material washed from the column with 2.5 mM phosphoric acid; (b) material eluted from the column with 5% acetone/95% 2.5 mM phosphoric acid; (c) material eluted from the column with 15% acetone/85% 2.5 mM phosphoric acid; (d) material eluted from the column with 50% acetone/50% 2.5 mM phosphoric acid.

175

and d show the material eluted from the column with successively higher concentrations of acetone. Fig. 3a-d show the corresponding data for urine acidified before loading onto the column and eluted with acidified solvents. Without acidification, ap- preciable amounts of organic material were not retained on the column, and were present in the urine passed through the column (data not shown) or in the water used to wash the column free of urine prior to acetone elution (Fig. 2a). In contrast, little organic material was unretained when urine and wash water were acidified (Fig. 3a).

Based on analytical HPLC with UV detection, increasing the amount of XAD-2 resin is effective in increasing the amount of organic material retained from a 500-ml sample of urine. Reversed- phase HPLC at neutral pH appears to have approximately the same order of magnitude of recovery as large XAD-2 columns (Fig. lc versus Fig. 2b-d), while reversed-phase HPLC at acidic pH yielded the best recoveries of organic material (Fig. 3b-d).

Comparison of the genotoxicity of urine material retained by the XAD-2 resin and preparative HPLC procedures

Extracts were prepared from 3 urine samples using different column procedures, and were tested for chromosome-damaging ability using CHO cells (Table I). Extracts were individually adjusted in concentration prior to testing to normalize them

on the basis of the creatinine content of the original urine specimens. Sample 'TS' is the same urine sample characterized by analytical HPLC in Figs. 1-3, while sample 'TS II ' is a second sample from this heavy smoker and sample 'PH ' is from a moderate smoker and coffee drinker.

After chromatography on the small XAD-2 column, the extracts from these 3 urine samples showed no significant clastogenic activity at the highest dose tested (4 mg /ml creatinine equivalence), while the large-scale XAD-2 column yielded extracts which were clastogenic in the case of all 3 samples. Preparative HPLC of sample 'TS' at neutral pH yielded 3 extracts with clastogenic activity. Had the material on the HPLC column not been separated into 3 fractions, but eluted in one fraction by a high concentration of acetone, this single fraction would have probably been more clastogenic than that isolated from the same volume of urine by the large XAD-2 column.

For all 3 urine samples examined, substantially more genotoxic material was obtained by reversed-phase HPLC utilizing acidified urine and elution solvents than by either large-scale XAD-2 column or preparative reversed-phase HPLC at neutral pH (Table 1).

Material eluted from the HPLC column by 2.5 mM phosphoric acid (column wash prior to acetone elution) was not completely soluble at creatinine equivalences of 4.0 and 2.0 mg/ml. Both of these doses caused mitotic inhibition in CHO cells, while

TABLE 1

COMPARISON OF THE CLASTOGENIC ACTIVITY OF URINE SAMPLES CONCENTRATED BY XAD-2 RESIN COL- UMNS AND PREPARATIVE HPLC PROCEDURES

Individual a Concentration procedure and fraction b

Small Large Prep HPLC; pH 7.0 Prep HPLC; pH 3.0 XAD XAD 1 2 3 1 2 3

CHO cells containing chromosome aberrations ¢ (%)

TS 1 12 6 12 19 11 20 50 TS II 1 27 - - - 17 12 41 PH 2 29 - - - 21 27 32

a TS is a heavy smoker (60 cigarettes/day). The two samples were taken several weeks apart. PH smokes 20-30 cigarettes and drinks 5 cups of coffee per day.

b Fraction 1: 5% acetone eluate; fraction 2: 15% acetone eluate; fraction 3: 50% acetone eluate. c Clastogenic activity is presented as the percentage of CHO cells containing at least one chromatid or chromosome aberration after

exposure to urine concentrates for 3 h and then cultured for 16 h, stained and analysed. Data displayed are for the highest dose: 4.0 mg/ml creatinine equivalence. The background level of chromosome aberrations is 1 or 2%.

176

at a dose of 1.0 mg/ml, no significant clastogenic activity was seen (data not shown).

Discussion

The identification of potential carcinogens and the elucidation of patterns of human exposure to known and unidentified carcinogens are both necessary prerequisites for intervention programmes designed to reduce or eliminate human exposure to carcinogens. Despite considerable effort, it has been difficult to obtain precise data on the exposure of individuals or population groups to carcinogens. One technique which holds some promise is the screening of human urine for genotoxic agents. Implementation of such screening requires adequate procedures for concentrating genotoxic materials from urine in a form suitable for application to short-term genotoxicity tests.

Aeschbacher and Chappuis (1981) reported that the urine of subjects consuming coffee does not contain elevated levels of mutagenic material. We carried out preliminary experiments on the clastogenic properties of urine from coffee drinkers and controls by freeze-drying whole urine, then dis- solving the residue in water and applying it to cells in tissue culture medium (Dunn, unpublished results). Our own experiments suggested an increased clastogenic activity in the urine of coffee drinkers but deficiencies in the methodology were apparent. Firstly, there was no clear basis on which to correct for the varying volume of urine excreted per hour by different subjects, or the body weight of the subjects (which differed in some case by a factor of 2 or more). Secondly, the large amount of salts and hydrophilic organic material in the urine samples strictly limited the degree to which urine samples could be concentrated prior to application to cells.

Creatinine is excreted in the urine at an approximately constant rate per hour per unit body mass. To correct for subject-to-subject differences in urine volume and body weight, we decided to normalize all results on the basis of the creatinine levels of the original urine samples. To allow concentration of genotoxic materials from urine, we decided to separate non-polar and moderately polar organic material in urine from salts and highly polar materials by using preparative re-

versed-phase HPLC. Because pH can affect the retention of some compounds on reversed-phase HPLC columns, chromatography was tried both at neutral pH and under mildly acidic conditions. Chromatography at alkaline pH was not at- tempted, as reversed-phase columns rapidly suffer irreversible damage at high pH.

The results of the current study indicate the need to critically appraise any extraction scheme used to concentrate organic material from urine for genotoxicity testing. Such an appraisal may be aided by chemical procedures such as analytical HPLC, designed to evaluate the efficiency of extraction schemes. Both analytical HPLC data and the results of chromosome aberration testing indicate that column chromatography of urine on small amounts of XAD-2 resin, while successfully used by a number of researchers, is inadequate in re- taining and concentrating more than a small amount of the organic material present in urine. The use of columns containing larger amounts of XAD-2 can increase the amount of organic material adsorbed, but involves considerable time and expense in the preparation of resin. Prepara- tive reversed-phase HPLC at acid pH is simply and rapidly performed, and appears to be superior to even large-scale XAD-2 columns in the retention of genotoxic material.

It should be noted that, while in the present experiments acidification appeared to be helpful in aiding the retention of both UV-absorbing material and genotoxic activity on the reversed-phase column, it may not be necessary or even desirable for all types of genotoxic materials in urine. Any basic compounds which are protonated at pH 3.0 but not at neutral pH would be expected to be retained better on the column without acidification.

The genotoxic material isolated from the preparative reversed-phase HPLC column was eluted in the form of 3 distinct fractions of differing hydrophobicity. This fractionation need not be done, and instead the material could be eluted in one fraction with the use of 50% acetone. How- ever, the separation of the organic material into 3 fractions may have considerable usefulness in un- derstanding the nature of the genotoxic material present in urine. At the very least, the presence of clastogenic activity in 3 chromatography fractions suggests that the genotoxicity of urine does not

arise from a single compound. The procedures developed here have been

utilized in an examination of lifestyle factors af- fecting the genotoxicity of urine (Dunn and Curtis, 1985). The urine of smokers and coffee drinkers contains elevated levels of genotoxic materials, which appear to be related to the presence of chemicals that can generate active oxygen species during incubation with cells.

Acknowledgements

This study was supported by grant No. 6610- 1329-52 from the National Health Research and Development Programme of Health and Welfare Canada. The Threshold Fund of Hampshire Col- lege, MA, is thanked for a travel grant to J.R. Curtis.

References

Aeschbacher, H.U., and C. Chappuis (1981) Non-mutagenicity of urine from coffee drinkers compared with that from cigarette smokers, Mutation Res., 89, 161-177.

Anderson, R.W., W.H. Puckett, W.J. Dana, T.V. Nguyen, J.C. Theiss and T.S. Matney (1982) Risk of handling injectable antineoplastic agents, Am. J. Hosp. Pharm., 39, 1881-1887.

Beek, B., I. Aranda and E. Thompson (1982) Induction of sister-chromatid exchanges, cell-cycle delay and chro- mosomal aberrations by human urine concentrations, Mu- tation Res., 92, 333-360.

Bos, R.P., A.O. Leenaars, J.L.G. Theuws and P.T. Henderson (1982) Mutagenicity of urine from nurses handling cyto- static drugs: influence of smoking, Int. Arch. Occup. En- viron. Health, 50, 359-369.

Bos, R.P., J.L.G. Theuws and P.T. Henderson (1983) Excretion of mutagens in human urine after passive smoking, Cancer Lett., 19, 85-90.

Dunn, B.P., and J.R. Curtis (1985) Clastogenic agents in the urine of coffee drinkers and cigarette smokers, Mutation Res., 147, 179-188.

177

Falck, K., M. Sorsa, H. Vainio and H. Kilpikari (1980) Muta- genicity in urine of workers in rubber industry, Mutation Res., 79, 45-52.

Gibson, J.F., P.W. Baxter, R.B. Hedworth-Whitty and D. Gompertz (1983) Urinary mutagenicity assays: a problem arising from the presence of histidine associated growth factors in XAD-2 prepared urine concentrates, with particu- lar reference to assays carried out using the bacterial fluctuation test, Carcinogenesis, 4, 1471-1476.

Guerrero, R.R., D.E. Rounds and T.C. Hall (1979) Bioassay procedure for the detection of mutagenic metabolites in human urine with the use of sister chromatid exchange analysis, J. Natl. Cancer Inst., 62, 805-809.

Iosefsohn, M. (1982) Creatinine in serum and urine with fuller's earth, in: W.R. Faulkner and S. Meites (Eds.), Selected Methods for the Small Clinical Laboratory, Vol. 9, Ameri- can Association for Clinical Chemistry, Washington, DC.

Laires, A., H. Borga, J. Rueff, M.I. Gomes and M. Halpern (1982) Urinary mutagenicity in occupational exposure to mineral oils and iron oxide particles, Carcinogenesis, 3, 1077-1079.

Legator, M.S., T.H. Conner and M. Stoeckel (1975) Detection of mutagenic activity of metronidazole and niridazole in body fluids of humans and mice, Science, 188, 1118-1119.

Minnich, V., M.E. Smith, D. Thompson and S. Kornfeld (1976) Detection of mutagenic activity in human urine using mutant strains of Salmonella typhimurium, Cancer, 38, 1253-1258.

Nguyen, T.V., J.C. Theiss and T.S. Matney (1982) Exposure of pharmacy personnel to mutagenic antineoplastic drugs, Cancer Res., 42, 4792-4796.

Putzrath, R.M., D. Langley and E. Eisenstadt (1981) Analysis of mutagenic activity in cigarette smoker's urine by high performance liquid chromatography, Mutation Res., 85, 97-108.

Stich, H.F., M.P. Rosin, C.H. Wu and W.D. Powrie (1981) The action of transition metals on the genotoxicity of simple phenols, phenolic acids, and cinnamic acids, Cancer Lett., 14, 251-260.

Yamasaki, E., and B.N. Ames (1977) Concentration of mutagens from urine by adsorption with the nonpolar resin XAD-2: cigarette smokers have mutagenic urine, Proc. Natl. Acad. Sci. (U.S.A.), 74, 3555-3559.

Documents

High-pressure liquid chromatography for isolation of clastogenic agents from urine