Original article

Dissection autoradiography

A screening technique using storage phosphor autoradiography

to detect the biodistribution of radiolabelled compounds

Catherine Walker, Clayton E. Walton, Julie M. Fraser, Fred Widmer, Xanthe E. Wells*

CSIRO Molecular Science, PO Box 184, North Ryde NSW 1670, Australia

Received 10 January 2001; accepted 8 August 2001

Abstract

Introduction: This study reports an alternative, rapid, whole body autoradiography technique which utilises storage-phosphor imaging

technology. Conventionally, tissue or whole body sections have been used to examine the distribution of radiolabelled test compounds.

However, the information acquired relates only to the sections examined, and the amount of radioactivity within the whole organ cannot be

quantified.We have developed a rapid semi-quantitative technique that produces a concise visual representation of the distribution of the isotope

throughout the entire animal: dissection autoradiography (DAR). Methods: By dissecting a mouse which has been administered 14C-labelled

methotrexate (MTX) and drying the tissues on a gel dryer, whole organs and aliquots of body fluids can be exposed to a phosphor imaging plate.

The data obtained was analysed with the software associated with the phosphor imaging system and, by using 14C standards, the amount of 14C

per total organ or tissue was quantified relative to other samples. Another widely used method to detect radiolabelled material in vivo is tissue

solubilisation (TS) followed by liquid scintillation counting (LSC). This conventional method was compared with DAR. Results: The new

technique described in this communication was found to have a high level of reproducibility (R2= 88–95%).Whilst DARwas less sensitive than

TS and LSC, trends over time in the biodistribution of 14C-MTX throughout most tissues were consistent between techniques. Discussion:

Whilst TS and LSC was a more sensitive technique, it was labour intensive and expensive in terms of consumables and time when compared

with DAR. Dissection autoradiography has the potential to be used to screen quickly large numbers of samples in the biodistribution studies of

various conjugates, isomers, derivatives or formulations of a parent compound, following a variety of routes of administration.D 2002 Elsevier

Science Inc. All rights reserved.

Keywords: Biodistribution; Dissection autoradiography; Methods; Mice; Whole body autoradiography

1. Introduction

Whole body autoradiography (WBAR) was developed in

the 1950s (Ullberg, 1954). A radiolabelled agent of interest

was administered in vivo, the animal euthanased at various

times thereafter, the carcass frozen and sagittal sections

freeze-dried, placed against an X-ray film-ray film, and

exposed. The technique enabled a detailed and complete

view of the distribution of the administered compound

throughout the entire body and has been used extensively

in pharmacokinetic and toxicological studies. However, con-

ventional X-ray film-ray film has a narrow dynamic range

and poor sensitivity to weak b-emitters like 14C and 3H, and

thus requires exposure for periods of up to several months.

Storage-phosphor autoradiography uses phosphor imag-

ing (PI) plates which have a fine coating of photostimulable

phosphors that detect and store ionising radiation. When hit

by a laser beam, this energy becomes luminescent and can be

digitised to reproduce an image of the radioactive specimen

(Kanekal, Sahai, Jones, & Brown, 1995). Phosphor imaging

plates have a linear dynamic range of up to five orders of

magnitude and the intensity of the luminescence is propor-

tional to the intensity of the original radiation. The data,

expressed in pixels, can be analysed for each delineated area.

This technology permits examination of both high and low

levels of radiation simultaneously and markedly reduces the

period of exposure of the sample, to hours instead of weeks

(Labarre, Papon, Moreau, Madelmont, & Veyre, 1998).

1056-8719/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved.

PII: S1056 -8719 (01 )00156 -3

* Corresponding author. Tel.: +61-2-9490-5060; fax: +61-2-9490-5020.

E-mail address: xanthe.wells@csiro.au (X.E. Wells).

Journal of Pharmacological and Toxicological Methods 45 (2001) 241–246

Whilst the use of PI plates reduces the exposure time,

conventional methods of tissue preparation for WBAR

have some shortcomings. Sectioning and dehydrating/

freeze- drying sample is very labour intensive and time

consuming. The image generated pertains to that section

alone, rather than the whole body, so that multiple sections

must be examined to ensure that all body tissues have

been included. When the amount of radioactivity is low in

a given tissue or the tissue sample or organ is tiny, the

image may not be visible within the section, particularly

when the isotope used is a weak emitter. The precise

location of radioactive material within for example various

anatomical regions of the gastrointestinal tract is difficult,

and requires tissue fixation, staining, and histological

expertise. We have developed a dissection autoradiography

technique (DAR) which examines entire organs, tissues or

aliquots of body fluids, enabling rapid preparation and

detailed analysis. Potentially, large numbers of individuals

can be examined, with a turn around time for each animal

of as little as 24 h.

Quantification of autoradiographic data has always been

problematic. Quantitative densitometry (Cross, Groves, &

Hesselbo, 1974) and later digital whole body image ana-

lysis (d’Argy, Sperber, Larsson, & Ullberg, 1990; Zane,

O’Buck, Walter, Robertson, & Tripp, 1997) utilises the

mathematical relationship between autoradiographic image

intensity and radioactivity detected using X-ray films (Irons

& Gross, 1981). However, direct measurement of radio-

activity by Geiger Muller counter following tissue solubi-

lisation (TS) or flask combustion of tissues has been

considered superior (Benard, Burgat, & Rico, 1985). In

phosphor imager systems, the computerised images gener-

ated from exposing the radioactive samples to PI plates can

be analysed using the software connected to the system. In

our DAR method, we measured the number of pixels in

each tissue or organ and converted them to counts using a

set of radioactive standards. Thus we could compare the

relative radioactivity within and among individual samples,

enabling semi-quantitative analysis. Our method was com-

pared with the results obtained by direct measurement of

samples from the same animal using TS and liquid scintil-

lation counting (LSC).

2. Materials and methods

2.1. Radiochemicals

14C-methotrexate (MTX) was prepared in house. Briefly,14C-glutamic acid was added to activated APA (4-[N-2,4-

diamino-6-pteridinylmethyl)-N-methylamino]benzoic acid;

Sigma Chemical Co.). Purification of 14C-MTX was per-

formed using thin layer chromatography. The precise

amount of 14C injected into the mice was determined by

LSC as described below.

2.2. Animals

Mature female BALB/c mice, with a mean weight of

22.5 ± 1.7 g, were bred on site. Approval for the animal

experimentation was obtained from the CSIRO Molecular

Science Animal Ethics Committee.

2.3. Animal treatments

Mice were administered 14C-MTX in 50ml of MTX 0.1%

in polyethylene glycol/phosphate buffered saline 1:1 via

injection into the tail vein. Following dosing, mice were

allowed free access to food and water. Urine and faeces

were collected during the post-administration period, after

which the mice were euthanased at various time intervals by

an intraperitoneal injection of pentobarbitone sodium (Nem-

butal1, Abbot Laboratories, Asquith, NSW, Australia).

Mice were washed to remove traces of radioactive material

from the feet and tail that may have been acquired from

faeces or urine on the cage floor.

2.4. Preparation of samples

The body was opened via a midline incision and

dissected organs (Table 1) weighed, laid out on a stack

of four sheets of chromatography paper (Advantec MFS

Inc, Pleasanton, CA, USA, Grade: No. 1514A), and

labelled by writing on the paper. The gastrointestinal tract

was stretched out and the mesentery removed. The stom-

ach, duodenum, jejunum/ileum, caecum, and large colon/

rectum were identified and each section tied off with

cotton thread. Urine, where available, was collected via

cystocentesis. The skin and brain were removed from the

carcass, which was rolled out and crushed with a heavy

glass rod, prior to being placed on the paper. The skin was

wet with ethanol and placed fur-side down on the paper. A

slurry of faeces was made in water, the total weighed, and

a proportion placed on the paper, together with aliquots of

urine. The paper stack was covered with plastic film and

placed on a gel dryer (Hoefer Scientific Instruments, San

Francisco, CA, USA), paper down, under vacuum for at

least 4 h at 80�C.

2.5. Autoradiography

The top two sheets, together with 14C microscales

(Amersham, Buckinghamshire, UK) were each covered with

plastic film and exposed to a PI plate (Molecular Dynamics,

Sunnyvale, CA, USA) overnight at room temperature. The

plate was scanned on the phosphor imager (Molecular

Dynamics PhosphorImager 400E). The image from each

organ—tissue or aliquot of body fluid— together with the

microscales, was delineated by drawing a line around them

and labelled, the number of pixels within that area calcu-

lated (ImageQuant Software Version 3.3, Molecular

C. Walker et al. / Journal of Pharmacological and Toxicological Methods 45 (2001) 241–246242

Dynamics), and the data from each of the two sheets of

paper transferred to a spreadsheet (Microsoft1 Excel 97).

The amount of 14C in the microscales as described by the

manufacturer was used to calculate the value, in nCi, of each

pixel. Thus the amount of 14C in each of the outlined areas

was derived by multiplying the number of pixels within that

area by the radioactivity represented by each pixel. By

incorporating the amount of 14C in the administered dose

(as determined by liquid scintillation counting of the injec-

tate), the values were standardised to those of an adminis-

tered dose of 1mCi/mouse. The spreadsheet automatically

calculated the standardised amount of 14C per total organ or

tissue, or per gram of tissue. This data was then available for

analysis between or among test groups.

2.6. Reproducibility of results

Mice were injected with 1mCi 14C-MTX as described

above. The results obtained from several mice at various

time points were compared using regression analysis (Mini-

tab1 for Windows, version 10.2, Minitab Inc, PA, USA).

2.7. Comparison of this technique with tissue solubilisation

To compare the counts detected in various organs and

tissues by DAR with TS, followed by LSC, mice were

injected with 1.06mCi 14C-MTX, anaesthetized, and tissues

dissected as described above. However, weighed portions

of the liver, kidneys, lungs, spleen, heart, thyroids, mes-

entery, uterus, brain, and aliquots of urine and faeces were

frozen at � 20�C, with the remainder being placed on the

paper and processed as before. The frozen portions of

tissue were minced finely with scissors, and weighed

aliquots (50–100 mg) were solubilised (Soluene1� 350,

Packard, Groningen, The Netherlands) at 50�C. Samples

were decolourised by the addition of 2 � 0.1 ml of 30%

hydrogen peroxide, with swirling between additions. Scin-

tillation fluid (Hionic–FluorTM, Packard) was added and

samples examined in a scintillation counter (Packard Tri–

Carb 2100TR Liquid Scintillation Counter, Packard, Mer-

iden, CT, USA). Hionic–Fluor scintillant was used to

minimise chemiluminescence. Counts/gram of tissue ana-

lysed by each technique were standardised to an adminis-

tered dose of 1mCi 14C-MTX and converted into counts per

organ or tissue.

3. Results

3.1. Images from dissection autoradiography

The autoradiogram obtained when a mouse was injected

with 5mCi 14C-MTX into the tail vein and euthanased 1 h

later is presented in Fig. 1. It clearly demonstrates the

Table 1

Amount of 14C (nCi, mean, range) from the organs and tissues of mice injected intravenously with 50 ml of MTX 0.1% in polyethylene glycol/phosphate

buffered saline 1:1 containing approximately 1 mCi 14C-MTX and determined using dissection autoradiography

Time post-injection

1 h (n= 2) 3 h (n= 3) 6 h (n= 2) 24 h (n= 2)

Liver 36.1, 36.1–36.2 16.8, 9.6–29.9 10.6, 9.7–11.4 6.7, 6.5–6.9

Kidneys 6.3, 6.3–6.3 3.0, 2.4–4.0 3.7, 3.6–3.7 2.4, 2.0–2.8

Stomach 2.1, 2.0–2.2 4.9, 2.8–7.5 3.9, 3.9–3.9 0.5, 0–1.0

Small intestinea 66.3, 65.2–67.3 30.0, 18.6–50.4 16.6, 15.3–17.9 6.9, 5.4–8.3

Large intestineb 5.2, 5.1–5.3 25.8, 23.6–27.4 18.4, 15.5–21.2 4.6, 3.4–5.3

GIT c 73.5, 72.3–74.8 60.2, 46.1–85.3 38.8, 34.6–42.9 12.0, 10.3–13.6

Faecesd 0, 0–0 8.2, 0–24.6 62.0, 42.0–82.2 72.5, 61.6–83.4

Spleen 0.7, 0.7–0.8 0.5, 0.5–0.6 0.8, 0.6–1.1 0.5, 0.4–0.7

Heart 0.4, 0.4–0.5 0.3, 0.2–0.3 0.4, 0.3–0.4 0.1, 0–0.1

Lungs 1.3, 1.2–1.4 0.7, 0.6–0.8 0.8, 0.8–0.9 0.2, 0.–0.5

Adrenals 0.08, 0–0.16 0.07, 0–0.12 0.05, 0–0.09 0.04, 0–0.07

Thyroids 0.5, 0.3–0.6 0.6, 0.5–0.6 0.5, 0.5–0.6 0.2, 0–0.3

Thymus 0.59, 0.35–0.83 0.27, 0.18–0.33 0.40, 0.27–0.53 0.09, 0–0.17

Mesenteric LNe 0.6, 0.5–0.7 0.2, 0–0.3 0.3, 0.3–0.4 0.1, 0–0.1

Other LNf 0.14, 0–0.27 0.24, 0.15–0.40 0.18, 0.12–0.23 0.06, 0–0.12

Uterus and bladder 8.4, 2.7–14.1 1.6, 1.4–1.8 1.7, 1.1–2.3 0.6, 0–1.1

Carcass 15.3, 14.4–16.2 11.8, 10.8–13.3 12.5, 11.5–13.5 3.7, 0–7.4

Skin 44.5, 39.3–49.7 24.4, 19.0–34.1 33.2, 31.9–34.6 15.0, 11.8–18.3

Fatg 1.5, 1.4–1.5 1.5, 1.0–2.3 1.1, 1.0–1.1 0, 0–0

Results have been standardised to an injected dose of 1mCi/mouse.a duodenum + jejunum;b caecum + colon + rectum;c small + large intestine;d total counts in collected faeces;e lymph node;f axillary + brachial LNs;g inguinal fatpad.

C. Walker et al. / Journal of Pharmacological and Toxicological Methods 45 (2001) 241–246 243

distribution of 14C throughout the entire mouse, graph-

ically delineating the relative radioactivity in each of the

dissected organs and tissues.

3.2. Reproducibility of results

When the data were compared overall for all the tissues

from each mouse (Table 1), the coefficients of determination

(R2) were 99% for 1 h, 88% for 3 h, 95% for 6 h, and 98%

for 24 h post-injection. This indicated a strong correlation

among samples, regardless of the time interval between

administration of the test compound and sample collection.

3.3. Comparison of this technique with tissue solubilisation

The amount of radioactivity detected tended to be higher

from TS than DAR (Table 2), indicating that the former

technique was more sensitive. This is further evident in the

Fig. 1. Dissected autoradiography image obtained from injecting a BALB/c mouse intravenously with 5mCi 14C-MTX 1 h previously.

C. Walker et al. / Journal of Pharmacological and Toxicological Methods 45 (2001) 241–246244

finding that radioactivity could be detected by DAR in only

one brain sample, whereas it was detected in all brains

using TS.

In some organs, for example liver, there was a marked

disparity in values observed between the two methods. This

is due, in part, to three factors related to the nature of 14C, a

relatively weak b-emitter with a maximum penetration of

only 0.04 mm through tissue sections (Kanekal et al., 1995).

Firstly, the larger the tissue being examined, the thicker the

gel-dried sample on the paper used for DAR. Consequently

there is quenching of the radiation from deep within the

sample and the penetration of 14C to the PI plate will vary

from the surface of the sample to that from within. Secondly,

the plastic film used to cover the sample before it is placed on

the PI plate absorbs some of the radioactivity (Laskey, 1990).

Thirdly, where the tissue is relatively wet, or a liquid for

example urine, and is placed on the filter paper, some of

the material is aspirated through the layers by the action of

the gel dryer. Therefore, radioactivity is present through the

depth of each layer of paper and would also be subject to

quenching. In this system, radioactivity is only examined on

two of the four layers of paper used, which reduces the

number of PI plates used per mouse. Whilst TS can be

subject to chemical and colour quenching, penetration of the

radiation is not an issue in its detection. However, for most

tissues the trends over time were reasonably consistent

between TS and DAR, indicating that DAR could be used

to measure the time-dependent biodistribution of a radio-

labelled compound.

The variation in readings between the two methods could

also have been contributed to by the errors introduced in both

systems with weighing small amounts (< 100 mg) of tissue

and then extrapolating the total radioactivity of an organ

based on the value obtained per gram of tissue. This weigh-

ing error is not a factor when organs and tissues are analysed

whole, as described in our method. Whilst multiple samples

were analysed to minimise the error, it may contribute to

variation in readings obtained with TS.

4. Discussion

The use of PI plates instead of X-ray film has reduced the

exposure time necessary to detect radioactivity, and the

software connected with the system permitted easy quan-

tification. Conventional whole body autoradiography exam-

ines sections from animals administered radiolabelled

compounds and enables visualisation of their distribution

throughout the entire carcass. However, the preparation of

these samples is time consuming— the animal must be deep

frozen with or without embedding, then sectioned whole, the

sections manipulated onto microscope slides or adhesive

tape and dehydrated before exposure to the PI plate, all of

which demand a high level of skill. Multiple sections must be

made and examined to encompass as many tissues and

organs as possible. However, only a proportion of the total

tissue is available for analysis.

We have developed a technique that utilises the advan-

tages of PI technology yet reduces the sample preparation

time. It took less than 1 h from the time of euthanasia to

placing the samples on the gel dryer (which can be run

overnight if necessary). Analysis of samples once they have

been exposed to the PI plate took approximately 20 min.

Consequently, a single operator could process many animals

per day permitting the rapid screening of large experimental

numbers. The rate limiting steps were the availability of gel

dryers and PI plates. Expensive consumables used in TS and

LSC, for example solubiliser, scintillant, and scintillation

vials, were not needed and disposal of toxic waste other than

the isotope was not an issue.

The DAR technique generates data which showed an

acceptable level of reproducibility among samples and could

be used for relative quantitation between different time

points compounds or routes. DAR facilitated the rapid and

comprehensive examination of a wide variety of samples,

particularly smaller tissues for example individual lymph

nodes, adrenal glands, and the thyroid. Thus the amount of

information obtained from each animal was increased, a

factor which has ethical implications.

Both WBAR of tissue sections and TS of tissue samples

are valuable techniques when detailed, accurate analysis is

required. Therefore, it is envisaged that this DAR method

will be an adjunct to existing approaches, and provide a clear

visual representation of the biodistribution of the test com-

pound. It could enable a simple, rapid screen of the biodis-

tribution of various conjugates, isomers, derivatives or

formulations of a parent compound. As well, alternative

routes of administration could be compared. It could be used

with other isotopes such as 35S or 125I, although, as with

conventional WBAR, 3H would be better detected using TS.

Table 2

Mean amount of 14C (nCi/organ) from mice injected intravenously with

50ml of MTX 0.1% in polyethylene glycol/phosphate buffered saline 1:1

containing 1.06mCi 14C-MTX

Time post-injection

1 h (n = 2) 3 h (n = 2) 6 h (n = 2) 24 h (n = 2)

DAR TS DAR TS DAR TS DAR TS

Liver 30.1 105.3 20.4 54.0 16.3 42.4 12.3 21.8

Kidneys 4.5 9.1 3.1 3.7 2.7 3.7 2.6 2.7

Lungs 1.2 1.6 0.3 1.3 0.9 0.9 1.0 0.6

Spleen 1.1 1.3 0.4 0.9 1.1 0.6 0.8 0.4

Heart 2.3 0.5 0.3 0.3 0.4 0.2 0.5 0.1

Thyroids 0.9 1.8 0.6 1.4 0.7 0.9 0.6 0.5

Mesentery 2.9 5.6 1.8 3.2 1.0 2.4 1.6 1.1

Uterus 2.2 3.6 1.2 2.0 1.6 0.8 0.9 0.4

Brain 0 0.7 0 0.6 1.5 0.4 0 0.3

Faecesa 0 02 9.8 28.82 35.7 38.7b 60.2 55.6

Urinec 1.7 8.8 1.82 5.52 0.1 0.3 0 0.004

Tissues were examined using both dissection autoradiography (DAR) and

tissue solubilisation (TS).a total counts in collected faeces;b n = 1;c nCi/ml urine collected by cystocentesis.

C. Walker et al. / Journal of Pharmacological and Toxicological Methods 45 (2001) 241–246 245

Acknowledgments

This work has been partly supported by F. H. Faulding &

Co Limited, Adelaide, SA, Australia. We thank Dr. Ross

Sparks, CSIRO MIS, Australia, for assistance with the

statistical analysis.

References

Benard, P., Burgat, V., & Rico, A. G. (1985). Application of whole-body

autoradiography in toxicology. Critical Reviews in Toxicology, 15,

181–215.

Cross, S. A. M., Groves, A. D., & Hesselbo, T. (1974). A quantitative

method for measuring radioactivity in tissues sectioned for whole body

autoradiography. Int J appl Radiat Isot, 25, 381–386.

d’Argy, R., Sperber, G. O., Larsson, B. S., & Ullberg, S. (1990). Computer-

assisted quantification and image processing of whole body autoradio-

grams. Journal of Pharmacological Methods, 24, 165–181.

Irons, R. D., & Gross, E. A. (1981). Standardization and calibration of

whole-body autoradiography for routine semiquantitative analysis of

the distribution of 14C-labeled compounds in animal tissues. Toxicology

and Applied Pharmacology, 59, 250–256.

Kanekal, S., Sahai, A., Jones, R. E., & Brown, D. (1995). Storage-phosphor

autoradiography - a rapid and highly sensitive method for spatial imag-

ing and quantitation of radioisotopes. Journal of Pharmacological and

Toxicological Methods, 33, 171–178.

Labarre, P., Papon, J., Moreau, M. F., Madelmont, J. C., & Veyre, A.

(1998). A new quantitative method to evaluate the biodistribution of

a radiolabelled tracer for melanoma using whole-body cryosectioning

and a gaseous detector: comparison with conventional tissue combus-

tion technology. European Journal of Nuclear Medicine, 25, 109–114.

Laskey, R. A. (1990). Radioisotope detection using X-ray film. In: R. J.

Slater (Ed.), Radioisotopes in Biology ( pp. 87–107). Oxford: Oxford

University Press.

Ullberg, S. (1954). Studies on the distribution and fate of S35- labelled

benzylpenicillin in the body. Acta Radiologica, (Suppl 118), 1–110.

Zane, P. A., O’Buck, A. J., Walter, R. E., Robertson, P., & Tripp, S. L.

(1997). Validation of procedures for quantitative whole-body autora-

diography using digital imaging. Journal of Pharmaceutical Sciences,

86, 733–738.

C. Walker et al. / Journal of Pharmacological and Toxicological Methods 45 (2001) 241–246246