The Uptake of ³⁸S in Cortical Bone

Experiments already reported (Kent, Jowsey, Steddon, Oliver, and Vaughan, 1956) on the uptake of ³⁵S given by intravenous injection to rabbits suggested that ³⁵S was taken up in two forms in cortical bone : (i) a labile form throughout the bone which is in large part removed by decalcify-ing agents; (ii) a non-labile form not removed by decalcifying agents. This was shown by chemical analysis to be chondroitin sulphate or a closely allied substance (Kent and others, 1956) and is probably associated with the ground substance of the bone matrix (Belanger, 1954; Vincent, 1954; Leblond, Bélanger, and Greulich, 1955).

The present study was undertaken in order to determine the effect of fixatives and embedding materials on the character of the ³⁵S incorporated in bone matrix as evidenced by reaction to hyaluro- nidase ;

the effect of hyaluronidase from different sources on ³⁵S incorporated in bone matrix ;

the exact location of non-labile sulphate in bone matrix at different times after injection.

Animals. The animals investigated were rabbits 6-8 weeks old, normal puppies 6-9 weeks old, and normal and rachitic puppies 26 weeks old. Rickets was produced as described in previous experiments (Jowsey, Sissons, and Vaughan, 1956) by the use of a rachitogenic diet devised by Mellanby (Mellanby, 1950). The animals were killed at various times after the injection of ³⁵S as Na₂³⁶SO₄ (2,000-3,000 p.c/kg), and pieces of cortical bone, 3 cm long, were taken from the tibiae.

Microradiographs were prepared by methods already described (Owen, Jowsey, and Vaughan, 1953 ; Jowsey, 1955) from sections from bones of all the animals in the experiment in order to check the degree of calcification present.

Autoradiographs were made from sections prepared by different methods described below. The section was attached to a slide with a solution of collodion in ether-alcohol. The film was floated on to the section from a bath of distilled water containing 10% glycerine at 28^o C. The films were thoroughly dried in a current of air from an electric fan and the sections stored in lighttight boxes in a refrigerator. The time of exposure varied according to the dose/kg the animal had received, and the time which had elapsed between injection and exposure. The autoradiographs were developed for 5 min in Digb (Kodak) at 18⁰ C, fixed in ‘hypo’ for 2-5 min, and washed well. In some cases after fixation the autoradiograph was floated off the section and remounted. Fast experimental film (Kodak Scientific Plate No. V 1001) was used in the preparation of autoradiographs from bone sections containing ³⁵S (Kent and others, 1956). The resolution given with this film is poor and there is much background fogging, but the uptake of ³⁵S in the skeleton is low. The maximum retention previously found in the femur 7 days after injection was 0-02% of the injected dose; therefore there is little ³⁵S present in any one section. The half-life of ³⁵S is only 87 days and if the sections have to be exposed twice it is essential to obtain a quick result. For purposes of comparison, photomicrographs have been included of autoradiographs of crosssections of the tibiae from rabbits and puppies in the same age groups used in other experiments when ⁹⁰Sr was given (Jowsey, Sissons, and Vaughan, 1956). ⁹⁰Sr is taken up in mineralized bone and not in osteoid (Vaughan, 1956).

Preliminary studies suggested that the effect of hyaluronidase on ³⁵S incorporated in bone matrix is influenced by methods of fixation and embedding. Conditions must be constant if consistent results are to be obtained. Crosssections were therefore prepared from pieces of cortical bone in the following ways:

(i) The tissue was fixed in absolute alcohol, embedded in methyl methacrylate monomer, and cross-sections were cut and ground to approximately 50 /z (Jowsey, 1955). The sections were immersed in chloroform for about 18 h to remove the monomer; they were hydrated, washed in distilled water, and then demineralized in EDTA for 24 h. The decalcified sections were then mounted.

(ii) The tissue was fixed in absolute alcohol for 3 days, during which the alcohol was changed 3 times. Pieces of bone were then decalcified in EDTA as previously described (Kent and others, 1956), and washed in running water for 1-2 h.

(a) The piece of bone was placed in ether-alcohol for 2 days ; the etheralcohol was changed once. The piece was then put in 1% collodion

Uptake of^3SS in Cortical Bone 371

solution (BDH) for 1 day, then in 4% collodion solution for 1 day, and then embedded in 6% collodion and allowed to set in a partial vacuum. When set, sections 5-to /z thick were cut from the blocks on an MSE microtome. Collodion was removed from the sections by immersion in ether-alcohol for 48 h. The sections were taken down through the alcohols to 70% alcohol and mounted.

(i) Other pieces of bone were dehydrated (70% alcohol for 2 h, 90% for 2 h, absolute for 3 changes of 4 h). They were then cleared in benzene (two changes of 4 h each), followed by three changes of increasing strengths of wax (melting-point 58^o C) mixed with benzene at 60^o C each for 2 h and then the block was cast with wax of the same meltingpoint. Sections were cut (5-10 /z) and mounted; the paraffin wax was removed from the sections with xylene and the sections were then taken down through the alcohols to 70% alcohol.

(iii) The tissue was fixed in neutralized formaldehyde-saline for 3 days and then embedded in

(a) collodion as described;

(b) paraffin as described.

(iv) The tissue was fixed in formaldehyde-saline at pH 9 and embedded in

(a) collodion;

(b) paraffin.

Some of these sections were washed extremely well and others less well before preparation of autoradiographs.

Some sections prepared in all the different ways described above after the first autoradiograph was floated off were treated for 6 h at 37^o C with o-i% testicular hyaluronidase (Evans Rondase) buffered at pH 5-6. The sections were then remounted and further autoradiographs made, a correction being allowed for the decay of ³⁵S in calculating exposure time. Other sections fixed in absolute alcohol and embedded in monomer and sections fixed in neutral formaldehyde-saline and embedded in paraffin were treated with staphylococcal hyaluronidase after the first autoradiographs were floated off. This hyaluronidase was buffered at pH 6 and incubated at 37^o C for 6 or 23 h. Each ampoule contained 160 turbidity-reducing units of enzyme, i.e. about 4 times as much as an ampoule of Evans Rondase. This was dissolved in 10 ml of o-2% gelatin containing 2X io^-4 M sodium pyrophosphate adjusted to pH 7 to guard against metal inactivation. Some sections treated with appropriate buffer solutions and without hyaluronidase were also used for the preparation of autoradiographs.

The effect of variation in fixation and embedding on the action of testicular hyaluronidase in removing ³⁵S from bone matrix are shown in table 1.

Testicular hyaluronidase removed ³⁵S from sections embedded in monomer or collodion but not from sections embedded in paraffin. The same results were obtained in material fixed in absolute alcohol, neutral formaldehyde-saline, or formaldehyde-saline at pH 9.

Testicular hyaluronidase removed all ³⁵S from both normal and rachitic osteoid except when the sections were embedded in paraffin. Staphylococcal hyaluronidase failed to remove ³⁵S from both paraffin and monomer embedded sections of normal bone. It was not used on rachitic bone. Buffer solutions at pH 5-6 and pH 7 alone were without effect on ³⁵S uptake.

Location of non-labile³⁵S

In animals killed 3 h after injection. Autoradiographs of cross-section of cortical bone from the normal puppy showed ³⁵S on the edge of the osteoid seam of building sites, in the area giving a positive PAS reaction in alcohol-fixed non-decalcified material (fig. 1, A, B). This pattern of uptake of ^3BS over the osteoid seam is markedly different from that seen for ⁹⁰Sr in a dog killed at the same time after injection when the ⁹⁰Sr appears in calcified bone behind the osteoid seam (fig. 1, c). ³⁵S was also found in the outer lamellae of periosteal and endosteal bone, but it was less easy here to be certain that it was present only on the extreme periphery of the seam since calcification in these sites is always more rapid than in internal osteones.

In animals killed 24 h after injection. There was rather more ³⁵S seen after decalcification than in the animal killed after 3 h. It was then present over the whole of the osteoid seam in active building sites (fig. 2, A) and in a thin line in the recently formed periosteal and endosteal bone. In the case of “Sr the radioactive isotope is again seen behind the osteoid border (fig. 2, B).

In animals killed 5 and 7 days after injection. The reaction in autoradiographs spreads back behind the osteoid seam into calcified bone (fig. 3, A, B). In periosteal and endosteal areas the reaction could be seen in a well-marked band with bone both internal and external to it showing no reaction. In the case of ⁹⁰Sr (fig. 3, c) there is a clear band of non-radioactive calcified bone between the osteoid seam and the bone containing ⁹⁰Sr.

Preliminary radiochemical estimations suggest there may well be quantitative differences in the uptake of ³⁵S in normal and rachitic osteoid. These studies are to be extended. Autoradiographs showed that ³⁵S was taken up in rachitic bone in the same sites as in normal bone, i.e. in areas where new osteoid tissue was being laid down.

It is not at present clear why, if tissue is embedded in paraffin, hyaluronidase fails to remove ^3BS. It is, however, important to recognize that this is so, and to follow a constant technique in histological study of ³⁵S uptake.

It has already been shown by chemical analysis (Kent and others, 1956) that the radioactive material extracted from decalcified bone after injection of ³⁵S is chondroitin ³⁵S sulphate or some closely allied substance. The fact that negative autoradiographs are obtained from decalcified sections of bone fixed in alcohol and embedded in monomer after treatment with testicular hyaluronidase, but not after staphylococcal hyaluronidase, supports the view that at least the greater part of the ³⁵S fraction is chondroitin sulphate A or C (Kent and Whitehouse, 1955 ; Meyer and Rapport, 1951 ; Meyer and others, 1956). Both chondroitin sulphate A and C are hydrolysed by testicular but not by bacterial hyaluronidase, while B is hydrolysed by both. Both A and C have been isolated from bone, though B has not (Meyer and others, 1956). Further work on the other characteristics such as optical rotation and solubilities of the material taking up ³⁵S will be required to discriminate between chondroitin A and C. It is possible, however, that there are in bone extremely small quantities of other substances which may be sulphated (Meyer and others, 1956). The ³⁵S present in the first lamella of the osteoid border of a building site probably is in a different chemical form to the ³⁵S present later in calcified tissue. In the former site the tissue is coloured a good pink with PAS in alcohol-fixed, undecalcified material, while 21 days later, or even 7 days later, the tissue containing the ³⁵S shows no distinctive PAS reaction, suggesting that ³⁵S is present in a different chemical combination. Leblond, Bélanger, and Greulich (1955) working with rats have shown a similar process occurring in periosteal bone, i.e. the deposition first of ³⁵S in non-mineralized osteoid followed by mineralization at a slightly older level of matrix. It has, indeed, been suggested by many workers that the chondroitin sulphate of the ground substances acts in some way as the means by which the mineral apatite crystals are associated with the collagen fibres (Kent and Whitehouse, 1955; Kent and others, 1956).

It should be noted that there are other areas in bone which give a positive PAS reaction but which do not take up ³⁵S. This may be due to metabolic inactivity of the tissue, which is probably the case in cartilage remnants or to the fact that the reacting material does not contain sulphur but one of the other mucoproteins known to occur in bone (Hismura, 1938; Eastoe and Eastoe, 1954; Glegg, Eidinger, and Leblond, 1954; Glegg and Eidinger, 1955)-

This work was begun on behalf of the Medical Research Council’s Committee on Protection against Ionizing Radiations. One of us (J. V.) has had a grant from the Medical Research Council for assistance and expenses. We are also indebted to the Nan Williams Fund for research into the Prevention and Cure of Leukaemia. We are indebted to Dr. Howard Rogers, National Institute for Medical Research, for the preparation and standardization of the staphylococcal hyaluronidase.