Relative distribution of mature erythrocytes, polychromatic erythrocytes (PE) and micronucleated PE on mouse bone-marrow smears: control observations

Mutation Research, 182 (1987) 203-209

203

Elsevier MTR08650

Relative distribution of mature erythrocytes, polychromatic erythrocytes (PE) and micronucleated PE on mouse bone-marrow smears: control observations E. Mirkova and J. Ashby Imperial Chemical Industries PLC, Central Toxicology Laboratory, Macclesfield~ Cheshire (Great Britain)

(Received24 November1986) (Revision received26 February 1987) (Accepted 11 March 1987)

Keywords: Bone-marrowsmears; Erythrocytes;Polychromaticerythrocytes;MicronucleatedPE; (Mouse); (Relative distribution). Summary Maps are presented showing the distribution of normal erythrocytes (NE), polychromatic erythrocytes (PE) and micronucleated PE (MPE) on 3 mouse bone-marrow smears. Adjacent areas of each slide were assessed along their whole length, and it was found that both the incidence of MPE and the ratio of PE : N E varied between different regions of the slide. These two variables were essentially independent of each other. The maps give the impression of clusters or islands of MPE on the slide, but these become less significant when the number of erythrocytes in those areas is taken into account. The present data confirm our previous conclusion that the resolving power of the bone-marrow micronucleus assay is related to the number of PE assessed for MPE per slide. Our data also illustrate that in instances where small departures from control values are observed, clarification/confirmation of the result can be obtained by extending the assessment of slides for MPE. It is proposed that the present recognition of the inherent sensitivity contained within this assay necessitates careful examination of the criteria for a positive response. The selection of PE suitable for assessment is discussed, and our recent 24-h oral corn oil control data are analysed.

In a previous paper (Ashby and Mohammed, 1986) we demonstrated that mouse bone-marrow slides were 'non-homogeneous' with respect to the incidences of micronucleated polychromatic eryihrocytes (MPE) among polychromatic erythrocytes (PE) when viewed over restricted

Correspondence; Dr. J. Ashby, Imperial Chemical Industries PLC, Central ToxicologyLaboratory, Alderley Park, Macclesfield, Cheshire SK10 4TJ (Great Britain).

fields ( < 1000 PE). In subsequent papers (Ashby and Mirkova, 1987a, b; Ashby and Ratpan, 1986) we demonstrated that the resolving power of this assay was substantially improved when the number of PE assessed per slide (animal) was increased from 300-500 to 2000-10000. Within this context it became necessary to establish, unequivocally, the local disposition of cells on bone-marrow slides. To achieve this we have constructed detailed maps of the position and relative distribution of PE, MPE and normal

0165-1161/87/$03.50 © 1987 ElsevierSciencePublishers B.V. (BiomedicalDivision)

204

MPE

~ g o

6ooI

~

~oo

~o ~• o ~

200

incidence

11 ( 12/1091 )

I

~

in selected

fields

16

43 (3 / 694)

4

(2 / 1289)

(4 ! 1012)

~._

I

i

500

40o

. -"

,oo

4536 cells at,sessed. overall ratio ol PE / NE O 85, overakl incrdence of MPE, 4 6 / 10no PE

" -- -- ~

0

:

I

i

:

I

•

i

=

I

i

i

I

I

I

:

I

i

i

I

J

J

Vernier units MPE

86 ( 11 / 1281 )

in s e l e c t e d fields

incidence

63 (8 / 1278)

1 ( 1 / 994)

44 (4 / 898)

32 (3 / 932 )

33 (5 / 1510)

700 800

~

500

~_~

400 /(/~\\

~ 300 o~

~

/

200

N E/,//pE /

\

,2"

~

l

7305 ceils assessed,

~

100

......

~,vnecT:;!!ac~2:i:ffMlapEE'N4E1,0000 P E

0 c

-

-

o

•

•

• '':

.

.

.

•

.

•

.:

•

•

• 0"~0

•

O--

I 0

.

.7 = t

J

I

1

I

~

J

J

I

MPE

incidence

I

i

in selected

i

I

i

126 ( 11 / 876)

93 ( 11 / 1188)

500

~ ~'~~ o~ o ¢~ Z o-

400

,o,3 -

*\, k

r Z-

3779 cells assessed,

/

..... , ....... ,E,NE 092: -

° incidence v e rof MPE, a l 9l5 / 1 0 0 0 PE

i

76 (4 / 522)

600

~,g

J

tields

~1 ( 11 / 998)

7OO

L

units

Vernier

",',

NE. ' r " /i //~pE//

///~\\

\ ~

300

/

200

\

//

100

0 • I

• L

.

• I

I

I

. I

. I

•

.

c

.

I Vernier

; I

I

--_-: ~ 2

: I

I

1

==-- . I

a i

= = i

• i

units

Fig. 1. Distribution of PE, N E and MPE across a local region of 3 slides. The lower section of each representation shows, as horizontal lines, the field traversed along a slide (against the horizontal vernier scale). These parallel lines represent exactly adjacent fields of vision. The length of a line was determined by the availability of cells for scrutiny. These lines therefore represent a visual representation of the distribution of MPE on the slide in the area studied (O, an MPE with perfect cell morphology; O, an MPE with a confirmed rnicronucleus within it, but which has slightly imperfect cell morphology). The upper curves display the distribution of PE and N E across the same area of the slide. These curves represent the result of joining the points which represent the number of cells (PE or NE) scored in the previous vernier division for all the sweeps of that part of the slide. The overall MPE incidence for each slide is shown in the box, and that within selected vernier regions, across the top of the representation (selected to the nearest 1000 PE). All calculations of MPE incidence shown are based on our perception of an erythrocyte suitable for scoring. Thus, the open-circled MPE shown were included, but may not have been by some investigators. Such investigators would also have reduced scores for PE and NE, thus the incidence of MPE may not be altered. The cumulative MPE incidences (re-calculated after each 500 PE) for these 3 slides are shown in Table 1.

205 erythrocytes (NE) on two control slides and one slide derived from a positive control animal. In addition, our recent 'historical' control data are analysed to show the effect of increasing the sampiing size when assessing slides for MPE. Methods

Three slides from male C 5 7 / B L / 6 J mice, consisting of two corn oil controls (slides 1 and 2) and a positive slide (slide 3), were employed for mapping. The two control slides were the highest scoring slides we have in our historical database for this strain of mouse. High values were selected to enhance the visualization of MPE distribution. In addition, our recent historical data for male CBA mice control slides are discussed. The slides were prepared according to a standardized procedure based on that of Schmid (1975). A femur was removed 24 h after dosing the animal and aspirated with fetal calf serum (FCS; 5 ml). The aspirate was mixed by gentle pipetting (30 × ) using a Pasteur pipette, and the suspension filtered through bolting cloth (150 #m; Henry Simon Ltd., Stockport, Great Britain). The resultant suspension was centrifuged for 5 rain (1000 rpm) and the pellet resuspended in the residual FCS ( - 0.1 ml) via pipetting (10 × ). This suspension was prepared as a glass-drawn smear according to Schmid (1975). Slides were air-dried (30 rain) and stained using a Hematech staining machine (polyhrome methylene blue/eosin). The slides were mounted and assessed for MPE in a standardized manner by reading the length of a slide, in parallel fields (1 field of vision apart) starting from the label end. A record was maintained of the PE number at which MPE were observed and the vernier readings noted for each MPE. The numbers of PE and NE between each vernier unit were also recorded. Results and discussion

In an earlier paper (Ashby and Mohammed, 1986) we presented 'rate of accumulation' graphs for MPE among PE on mouse bone-marrow micronucleus slides. There it was clear that the distribution of MPE across a slide was essentially homogenoeus when viewed over several thousand

PE, and this has recently been confirmed statisticaUy (Albanese, 1987). However, when viewed over fields of < 1000 PE, marked discontinuities are evident (cf. Fig. 1). Upon increasing the number of PE assessed from a few hundred to 2000-3000, the mean incidence of MPE stabilizes significantly (Table 1), and this forms the basis of our current practice of assessing at least 2000 PE for MPE in this assay. Nonetheless, the choice of 2000 PE is essentially empirical, despite the facts that it is consistent with the suggestion made by de Mitchell and Brice (1986) and is supported by our recent observations for D M H and M N N G in this assay (Ashby and Mirkova, 1987a, b). For routine purposes it may be sufficient to commence with a more limited assessment of slides, especially in cases where large group sizes of animals are employed. In cases where a weak positive response is indicated the extended assessment of slides would place the observations on a firmer basis; the precise number of PE assessed being dependent upon the required assay sensitivity (see later). Although the erythrocyte maps (Fig. 1) give the impression of clustering of MPE on some parts of the slides, these occur predominantly in areas rich in erythrocytes, thus the incidence of MPE is not subject to clustering. However, the relative distribution across slides of PE and normal erythrocytes varies more than we had expected (Fig. 2), and it is concluded that if the ratio of P E : N E is to be determined to provide possible evidence of chemically mediated bone-marrow toxicity, then between 1500 and 2000 erythrocytes should be counted to provide an accurate value for individual slides. In our recent studies (from which the present slides were taken) we have taken extreme care to prepare homogeneous smears. It seems unlikely that greater precautions to ensure complete cell mixing could be taken, so it is reasonable to assume that the present observations are of general validity. It follows that the majority of the variability recorded for micronucleus assay data in cases where < 1000 PE have been assessed is due to the under-sampling of slides, and this must, of necessity, reduce the resolving power of the assay. This problem is illustrated by recent data of Lin and Brusick (1986) where up to 7-fold changes of MPE incidences were reported between control

206 TABLE 1 C U M U L A T I V E I N C I D E N C E OF MPE ACROSS T H E 3 SLIDES SHOWN IN Fig. 1 WITH I N C R E A S I N G N U M B E R OF PE ASSESSED The accumulation was derived from the primary data, in the order in which the cells were assessed. It is evident that the final (overall) incidence is only apparent after between 1500 and 2000 PE have been assessed. In particular, a misleading incidence is derived after the assessment of only 500 PE, as still practiced by some investigators (see text). Slide No.

MPE/1000 PE at intervals of 500 PE (accumulated)

(cf. Fig. 1)

500

1000

1500

2000

2500

3000

3500

Ovrall incidence based on full assessment

1 2 3

10 8 10

6 6 13

6.7 5.3 9.3

5.5 4.5 8

7.5 5.6 8.8

5.7 5 10

5.1 4.3 9.7

4.6 4.4 9.5

1.2

1.1 |

1.0

•

0.9

•

0.8

• • •

•

/

•

P

•

• •

0.7

PE NE

0.6 0.5

SLIDE 3 1

I SLIDE 2 I

I SLIDE 1 I 0.4 0.3 0.2 0.1 0

B

I 500

I I I .... I 1000 1500 3000 4536 PE/NE

500

I ..... I 1000 1500 3000 7305

I 500

I 1000

I I .... I 1500 3000 3779

ratio for increasing sample size of PE

Fig. 2. Ratio of P E / N E , with increasing group size of PE assessed, for slides 1-3 (Fig. 1). The ratio was calculated, in each case, in the direction in which the slide was initially assessed for MPE. Thus, after counting the total number of PE and N E in the areas of the slide traversed we were able to calculate the overall P E / N E ratio shown on the right of each of the above figures. The rest of the numbers (ratios) shown in the figure were derived by taking the ratio for the first 3000 PE on each slide and then breaking this up into 2 x 1500 PE assessed, 3 x 1000 PE or 6 x 500 PE. It is clear that an acceptable approximation to the ratio obtained after 3000 PE is given by the 1000 PE values, but not by the 500 PE values.

207 a n d test g r o u p m e a n values for M P E incidences after assessing 500 P E p e r a n i m a l ; differences c o n s i d e r e d to b e non-significant. T h e recent ' h i s torical' c o r n oil c o n t r o l d a t a o b s e r v e d in this l a b o r a t o r y for m a l e C B A mice (as used in our recent e x p e r i m e n t s ) are shown in T a b l e 2. T w o p o i n t s are r e l e v a n t to the p r e s e n t discussion. First, the s t a n d a r d d e v i a t i o n (S.D.) associated with the g r o u p m e a n M P E incidences for b o t h oral a n d i.p. studies is r e d u c e d when c o m p a r i n g the 2000 P E assessment figure with the 1000 P E figure. Second, the i.p. m e a n c o n t r o l value a p p e a r s to be lower t h a n the c o r r e s p o n d i n g oral value. This r e d u c t i o n

in the S.D. of the m e a n M P E incidence has rec e n t l y led us to c o n c l u d e that the s p o n t a n e o u s i n c i d e n c e of M P E for i n d i v i d u a l strains of mice is p r o b a b l y a s s o c i a t e d with m i n i m a l variability, a p a r t f r o m t h a t i m p o s e d b y the u n d e r - s a m p l i n g of slides. A s a direct c o n s e q u e n c e of this we have recently t a k e n to e x t e n d i n g the assessment of the highest a n d lowest c o n t r o l slides in a study. This always results in the high value d r o p p i n g , a n d the low value rising t o w a r d s the historical m e a n (see T a b l e 2 a n d A s h b y a n d M i r k o v a , 1987b). This therefore h a s the effect o f r e d u c i n g the variance o f the c o n t r o l data, a n d t h e r e b y e n h a n c i n g the statistical

TABLE 2 CONTROL VALUES FOR THE INCIDENCE OF MPE ON SLIDES PREPARED FROM THE BONE MARROW OF MALE CBA MICE (n = 110) Selected slides from the oral control group which gave MPE incidences either above or below the historical mean after 2000 PE had been assessed were submited to an extended assessment (4000 PE). The high values were reduced and the low values increased, the net result was that the S.D.s of the group means were further reduced. The i.p. control animals had a significantly reduced incidence of MPE compared to the oral control animals (after 2000 PE, 1-sided Student's t test, p < 0.05), however, this difference was not apparent when only 1000 PE were assessed (not-significant by the same t test). For comparison, MPE incidences are shown after only 1000 PE had been assessed per slide, but slide assessment did not cease until 2000 PE had been evaluated. Slide values (1000, 2000 and 4000 PE) for each slide are shown in the same relative position of the 3 columns of this table. Route of adn~nisstriation

MPE/1000 PE Individual corn oil control animal data Number of PE assessed per slide (animal) 1000

2000

4000

6,4,3,5,6,5,7,7,3, 5,5,12,10,5,5,7,5,6,3, 5,9,3,5,6,7,7,5,

4,4,4,4,4.5,4.5,6,6,4, 5.5,5.5,6.5,8,4.5,4,4.5,4.5,5,5.5, 4.5,8,5.5,4,5,4.5,6,4

4,3,3,0,2,3,3,3,2,1, 3,0,5,3,4,5,1,3,0,4, 4,1,3,2,1,1,4,1,6,2, 2,3,1,5,3,3,2,1,4,5, 3,5,2,2,3,2,4,1,1,0, 2,0,5,0,2,5

3.5,2,3.5,1,1.5,2,2.5,2,2,1, 2,0.5,3,3,3.5,3,1.5,2.5,2,3, 3,1.5,2,2.5,3,3,3,2,3,2, 2.5,2,1,3,3,2.5,3,0.5,3,2.5, 3,2.5,3,2.5,2,2,3,1.5,1,2, 3,2,3,1,1.5,3.5

3.58±2.3(n=83)

3.2±1.57(n=83)

10,5,5,7,5,9,3,6,7, 3,3,3,2,1,3,0,1

8, 4.5, 4, 4.5, 4.5, 8, 5.5, 5, 6 (high) 2, 2.5, 2, 2, 1, 2, 0.5, 0.5 (low)

7 *, 3.2, 2.7, 3.5, 3.5, 5, 5.25, 3.5, 3.5, 3, 2.5, 2, 3.5, 2.2, 2, 1, 2.25 * [5.4 after 8000 PE assessed]

4.29±2.8(n=17)

3.68 + 2.37 (n = 17)

3.27 ± 1.43 (n = 17)

6,9,5,3,4,1,1,3,6,1, 3,3,3,2,5,2,3,1,3,5, 3,1,3,3,5,2,1

4.5, 4.5, 4, 2, 2.5, 2, 1, 2.5, 3, 0.5, 2, 2.5, 2;5, 1, 2.5, 3, 3, 3, 2.5, 3.5, 3, 1.5, 3, 1.5, 3, 1.5, 2

3.22±1.93(n=27)

2.5 +0.99 (n = 27)

>3.5MPE/I~OPE~=2~

<3.5MPE/IO~PE~=5~ Oral gavage

Extended readings for 17 of the above 83 controls (9 initially high, 8 initially low) i.p. injection

208 power of the assay. Another advantage of the adequate assessment of slides is that it enables one to become aware of genuine instances of responder variability in treated groups, a factor which has a major influence on the choice of statistical model (de Mitchell and Brice, 1986). It could be argued that the unit of this assay is the number of PE assessed per group, i.e., that conclusions derived from assessing 10 000 PE could be arrived at equally by reading 2000 PE from 5 animals or 500 PE from 20 animals. However, apart from the doubtful value of combining a series of approximate values, the extra use of animals necessitated by this alternative approach seems to be both unjustified and impractical. The low variability in control MPE incidences (Table 2) suggests that sole reliance might be placed upon historical control values, but that is not specifically recommended as each study should be appropriately controlled. Equally, however, greater use could be made of historical control values in statistical assessments of data. This concept could become increasingly important, because, as the resolving power of an assay increases, so also does the need to define clear criteria for a positive test response. For example, the difference in control MPE frequencies observed here for the oral and i.p. injection routes of exposure (Table 2) only achieved statistical significance after 2000 PE had been assessed. This effect may or may not be a real effect induced by corn oil, further data will be required to draw a conclusion; however, it focuses the need for adequate assay positive response criteria. One criterion for consideration is that the mean + S.D. of a treated group should be significantly different from that of the recent historical control data (for example, the last 100 animals). The additional requirements of reproducibility and a dose-related response would enhance further positive conclusions. Equally, it may be considered inappropriate to classify a weak response as seen here for corn oil as positive, and then it may be decided that the epithet of 'in vivo genotoxin' should be reserved for agents capable of inducing a mean response in treated animals of, for example, twice the historical mean. It is only when laboratories who routinely use this assay publish and discuss their control data that this question can be mutually resolved.

A topic worthy of general consideration is the uncertainty associated with the available definitions of the units of data for the bone-marrow micronucleus assay (MPE and PE) (cf. Racine and Matter, 1984; Heddle and Salamone, 1981; Heddle et al., 1983; and a discussion of these in Ashby and Mohammed, 1986). After extensive experimentation using different interpretations of these selection criteria, we have settled on the following. PE suitable for scoring should be distinct in hue from NE, adequately spaced from each other (but not be regarded as negligible if a minor overlap of cytoplasms exists) and to be of good, but not necessarily perfect, morphology. These same criteria apply to MPE except that they must also contain a body which is usually circular, always coplanar with the cytoplasm, > 0.1 of the diameter of the cell and capable of being discerned as red/red-blue in hue under bright illumination. If any doubts exist, the cell is simply added to the PE score. In some cases (as with procarbazine-induced MPE; Ashby and Mirkova, 1987a), large, slightly diffuse, irregular and paler staining micronuclei are produced. The importance of perfect cell morphology for scorable PE is emphasized in the criteria papers listed above, and if these are taken seriously, a large reduction in the number of PE and MPE considered suitable for scoring ensues. Thus, in Fig. 1 we have left as open circles those MPE which rigid morphological criteria would have eliminated from consideration (e.g., a slight discontinuity in the cell membrane etc.). However, we considered these as certain to be PE and certain to contain a micronucleus; thus they were included as MPE in our assessment. The issues raised in this paper become critical when assessing a possible weak positive test response. Similar concerns regarding the selection of cells for assessment and sample sizes were recorded by Galloway et al. (1986) when they compared the results derived by a group of independent cytogenetics who had assessed the same preparations of human lymphocytes for chromosomal aberrations.

Acknowledgements We wish to acknowledge the financial support of the European Science Foundation and the

209 European Medical Research Councils, and the e n c o u r a g e m e n t p r o v i d e d b y P r o f e s s o r F. K a l o janova of the Institute of Hygiene and Occupat i o n a l H e a l t h , T h e M e d i c a l A c a d e m y , Sofia.

References Albanese, R. (1987) The assessment of micronucleated polychromatic erythrocytes in rat bone-marrow, Mutation Res., submitted. Ashby, J., and E. Mirkova (1987a) Re-evaluation of 1,2-dimethylhydrazine in the mouse bone marrow micronucleus assay; observation of a positive response. En~'iron. Mutagen., 9, 177-181. Ashby, J., and E. Mirkova (1987b) The activity of MNNG in the mouse bone marrow micronucleus assay, Mutagenesis, in press. Ashby, J., and R. Mohammed (1986) Slide preparation and sampling as a major source of variability in the mouse micronucleus assay, Mutation Res., 164, 217-235. Ashby, J., and F. Ratpan (1986) Inactivity of glycerol formal in a mouse micronucleus assay: relationship to its teratogenicity, Environ. Mutagen., 8, 873-878.

de Mitchell, G., and A. Brice (1986) Investigations into parametric analysis of data from the in vivo micronucleus assay by comparison with non-parametric methods, Mutation Res., 159, 139-146. Galloway, S.M., P.K. Berry, W.W. Nichols, S.R. Wolman, K.A. Soper, P.D. Stolley and P. Archer (1986) Chromosome aberrations in individuals exposed to ethylene oxide, and in a large control population, Mutation Res., 170, 55-74. Heddle, J.A., and M.F. Salamone (1981) The micronucleus assay, I. In vivo, in: Eds, H.F. Stich and R.H.C. San (Eds.), Short-Term Tests for Chemical Carcinogens, Springer, Berlin, pp. 243-249. Heddle, J.A., M. Hite, B. Krikhart, K. Mavournin, J.T. MacGregor, G.W. Newell and M.F. Salamone (1983) The induction of micronuclei as a measure of genotoxicity, Mutation Res., 123, 61-118. Lin, G.H.Y., and D.J. Brusick (1986) Mutagenicity studies on FD and C Red No. 3, Mutagenesis, 1, 253-259. Racine, R.R., and B.E. Matter (1984) The micronucleus test as an indicator of mutagenic exposure, in: A.A. Ansari and F.J. de Serres (eds.), Single-Cell Mutation Monitoring Systems, Plenum, New York, pp. 217-232. Schmid, W. (1975) The micronucleus test, Mutation Res., 31, 9-15.