timing of sulphogalactolipid biosynthesis in the rat testis studied by ...

J. Cell Sci. 75, 329-338 (1985)

329

Printed in Great Britain © The Company of Biologists Limited 1985

TIMING OF SULPHOGALACTOLIPID BIOSYNTHESIS IN THE RAT TESTIS STUDIED BY TISSUE AUTORADIOGRAPHY CLIFFORD A. LINGWOOD Department of Biochemistry, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada and Departments of Clinical Biochemistry and Biochemistry, University of Toronto, Ontario, Canada

SUMMARY The testicular synthesis of sulphatoxygalactosylacylalkylglycerol (SGG) has been studied in the rat by autoradiography of frozen tissue sections following in vivo metabolic labelling. The results are consistent with the synthesis of this major mammalian germ-cell glycolipid at the zygotene and early pachytene stages of spermatogenesis. Further synthesis of SGG is prevented by the appearance of an inhibitor of galactolipid sulphotransferase activity at the mid-pachytene stage.

INTRODUCTION

Sulphatoxygalactosylacylalkylglycerol (SGG) is a major component of the mammalian male germ cell membrane (Murray et al. 1980). The biosynthesis of SGG has previously been demonstrated to be a marker of germ cell development in the rat. Measurement of the levels of SGG in the testes of rats during the first wave of spermatogenesis after birth have shown that synthesis of SGG occurs 15-20 days after birth, corresponding to the appearance of primary spermatocytes in the testis (Kornblatt et al. 1974). Once synthesized, SGG remains a stable component of the germ cell membrane throughout spermatogenesis, without turnover. Although SGG is synthesized over a very short period of the spermatogenic cycle, the exact cell responsible for this biosynthesis has not been demonstrated. Isolated pachytene spermatocytes or round spermatids show only a limited capacity for SGG synthesis in vitro (Letts et al. 1978), which cannot explain the biosynthesis in vivo. Thus, by inference, SGG biosynthesis has been ascribed to leptotene and zygotene primary spermatocytes. A method to show that this was indeed correct, would be tissue autoradiography after in vivo administration of 35SO4, since it has been shown that SGG is the only lipid labelled under these conditions (Lingwood, Hay & Schachter, 1981). However, autoradiographic demonstration of radiolabelled glycolipids in tissue samples is difficult since they may be removed during dehydration of the section. This is particularly true for SGG, which is fully saturated and therefore not fixed by osmium tetroxide. Mild fixation with formaldehyde, without dehydration, of frozen testicular sections, together with oral administration of isotope has now Key words: glycolipid, autoradiography, spermatogenesis.

330

C. A. Lingiwod

allowed the autoradiographic localization of SGG biosynthesis within the germ cell lineage.

MATERIALS AND METHODS

Autoradiography A sample containing 1 mCi of H235SO4 (43 Ci/mg New England Nuclear) was soaked into several pellets of rat chow. The radioactive food was given to an adult (300 g) Wistar rat and was eaten over an 18-h period. After this time, the rat was killed by asphyxiation with CO2 and the testes were removed and frozen in isopentane chilled in liquid nitrogen. Cryosections (5 jUm) of the frozen tissue were then cut. The sections were fixed for 10 min in neutral buffered formalin (3-7 % formaldehyde) and washed for 1 h in running tap water. The sections were rinsed four times with distilled water and dried at 37 °C. A coating of Nuclear Track Emulsion (NT-B-2 Kodak) maintained at 42°C was then applied in the dark. The slides were allowed to drain and then air dry for 2 h in the dark. The sections were sealed in a light-proof container and maintained at 4°C for 4 weeks. After this time, the slides were developed for 2 min in Dektol (Kodak). The reaction was stopped with 2 % acetic acid and slides were washed in three changes (1 min each) of Kodak fixer and for 30 min in running tap water. Sections were then stained with haematoxylin and eosin and mounted in aqueous mounting medium (glycerol gelatin, Sigma Chemical Co.). Some sections were observed without staining. The slides were observed under a Zeiss phase-contrast microscope and photographed using Ilford PAN F film.

Microdissection of seminiferous tubules Testes were removed from adult rats immediately following asphyxiation with CO2. Tunicae were removed and tubules dissociated mechanically into PBS (0-1 M-phosphate-buffered saline, pH 7-4) containing 1 % fructose. Lengths of tubule at different stages in the cycle of the seminiferous epithelium were dissected under transillumination using a fibre optic light according to the criteria of Parvinen (1982). The cycle was divided into three segments based on the most obvious of these criteria: (A) stages V I I - V I I I ; (B) stages I X - I ; (C) stages II—VI. Tubules were homogenized for assay of galactolipid sulphotransferase activity in vitro as described below.

Sulphotransferase assay Tubule segments were homogenized in a micro glass homogenizer in 0-32M-sucrose/l mMEDTA (pH7-4) to give a protein concentration of approximately 1-Omg/ml as measured by the Bradford (1976) method, using rabbit immunoglobulin as standard. Glycolipid substrate (25 nmol galactosylacylalkyl+glycerol), prepared as described previously (Lingwood, Sakac & Vella, 1983), was dissolved in 25p\ of 4 % Triton X-100 in chloroform/ methanol, 2:1 (v/v). The solution was taken to dryness under nitrogen and resuspended in O'l ml 0-lM-Tris-HCl (pH8-6) containing 10mM-ATP, 2-5mM-MgCl2, 80mM-K 2 SO 4 , 2jiM-3'phosphoadenosine 5'phosphosulphate (PAP35S) (2-3 Ci/^mol New England Nuclear) and 25 jl\ sulphotransferase preparation. The mixture was incubated at 37 °C for 2h and partitioned into chloroform/methanol 2:1 (v/v) and 0-88 % aqueous KC1. The lower phase was dried and subjected to thin-layer chromatography on silica gel G plates (chloroform/methanol/water, 65:25:4, by vol.) and 3 SGG was detected by autoradiography. Radioactive product was scraped from the plate, suspended in 4 ml ACS (Amersham) scintillation fluid and counted in an LKB Rackbeta II scintillation counter.

Fig. 1. Frozen sections of adult testis. A. Autoradiogram of unstained testis section following in vivo administration of [35S]sulphate. B. Similar section from untreated rat stained with haematoxylin/eosin. Arrows in A indicate heavily and lightly labelled tubules. X14.

331

B

332

C. A. Ltngwood

Assay for sulphotransferase inhibitor Microdissected tubule segments were homogenized as described above. The homogenate was centrifuged at 8000 # for 2 min. Inhibitory activity was assayed by including 50 /A of the supernatant in the in vitro sulphotransferase assay described above, using the homogenate from the testes of 20day-old rats as enzyme source (Lingwood, unpublished).

Mmst~K-"-

••-

:

&&*:*#

Fig. 2. Autoradiography of frozen testicular sections. Sections were stained with haematoxylin/eosin following autoradiography. Pictures have been printed to emphasize the silver grains. Each picture shows a section of two adjacent seminiferous tubules. Where possible, these have been assigned to a stage in the cycle: A: left, III; right, X I I - X I I I ; B: upper, XIII; lower, VII; C: left, VI; right, VIII; D: left, II; right, X; E: VII; F: left, X I I - X I I I ; G: IV. X87.

Testicular sulphogalactolipid biosynthesis

333

RESULTS

Autoradiography of frozen testicular sections after in vivo administration of H235SO4 shows that sulphation occurs primarily in the cells that are in or near the perimeter of the seminiferous tubules (Fig. 1A). Silver grains are detected in rings corresponding to the seminiferous tubules. Labelling is reduced in the interstitium and some tubules appear to be less intensively labelled (Fig. 1A, arrow). Using only the criteria of shape and position of elongating spermatids, presence of residual bodies and relative size of cells near the basal compartment, tubule sections were ascribed, where possible, to a particular stage in the cycle of the seminiferous epithelium. In order to quantitate the degree of labelling, a rectangular area 350^mX550^m (Fig. 2) from the basal lamina of the tubule towards the lumen was selected and silver grains were counted. At least three rectangles were counted for any given assigned tubule section. At least 500 grains were counted and, where possible, several sections at the same stage were measured. Representative sections are shown in

0! •

0-05

o-3

1

2

n-2 n - 5

3 4 5 6 7 8 9 10 11 12 13 14 Stage of seminferous epithelium

Fig. 3. Quantitation of labelling. Silver grains in assigned tubule stages (1-14 represent I-XIV) were counted as described in Materials and Methods. Value and standard deviation are shown.

334

C. A. Ungwood

Table 1. Regulation of galactolipid sulphotransferase activity in microdissected segments of seminiferous tubules Segment stages IX-I II-VI VII-VIII

In vitro sulphotransferase activity (pmol SGG synthesized/mg protein per h)

Sulphotransferase inhibitory activity* (% control incubation ± S.D.)

5-4 1-4 3-6

82-2 ±16-8 36-5 ±12-2 71-2±23-8

• Inhibitory activity was assayed as described in Materials and Methods by determining the effect of inclusion of a sample of the tubule homogenate supernatant in a standard control sulphotransferase assay.

Fig. 2. The difference in labelling intensity is most obvious in Fig. 2B where tubules at stages XIII and VII are in apposition. The quantitation of labelling throughout the seminiferous cycle is shown graphically in Fig. 3. Background values for silver grains in the lumen adjacent to each section have been subtracted. Maximum labelling occurs in stages XII-II, while the minimum is observed for stages IV-VIII. It can be seen, however, that labelling above background occurs at all stages. Treatment of the 35S-labelled testicular sections by traditional histological procedures involving dehydration and mounting in Bouin's fixative, results in the complete loss of silver grains from within the seminiferous tubule (Fig. 4). Silver grains within the interstitial compartment, however, are virtually unaffected. Assay of the in vitro galactolipid sulphotransferase activity in tubule segments following microdissection shows that maximum galactolipid sulphotransferase activity assayed in vitro is found for tubules at stages IX—I (Table 1), whilst maximal sulphotransferase inhibitory activity is found for stages II—VII (Table 1). Little or no inhibitory activity was found for segments from stages IX-I.

DISCUSSION

Testicular galactoglycerolipids cannot be fixed by osmium tetroxide and are therefore lost during the ethanol dehydration stage of fixation procedures used in preparing radiolabelled tissue sections for autoradiography. Mild fixation with formalin of frozen testicular sections avoids this loss and still retains a reasonable Fig. 4. Effect of fixation and dehydration on testicular autoradiograms. Adjacent tubules are shown and the interstitial cells are indicated (arrows), A. Autoradiogram of radiolabelled testicular section as described for Fig. 2. The majority of silver grains are found in the basal cells of the tubule and some in the interstitial cells. B. Autoradiogram of radiolabelled testicular section after dehydration in ethanol and treatment with Bouin's fixative. Essentially all grains within the seminiferous tubules have been removed, while labelling the interstitium is unaffected. Both sections are stained with haematoxylin and eosin. X225.

Testicular sulphogalactolipid biosynthesis

Fig. 4

SIS

336

C. A. Lingwood

degree of the morphology. The specific activity of labelling of testicular sulphoglycolipids has been increased by oral administration of [35S]sulphuric acid on food pellets, which provides a prolonged, though perhaps irregular, metabolic labelling period. Previous attempts at tissue autoradiography following intratesticular or intraperitoneal injection of'[ Sjsulphate were unsuccessful. The autoradiographic radiolabelling pattern (Fig. 1A) clearly demonstrates that the incorporation of [ Sjsulphate occurs in the cells closer to the basement membrane of the seminiferous tubule. It is not possible to define precisely the cells labelled with silver grains; however, the labelling pattern is consistent with SGG biosynthesis by primary spermatocytes (Fig. 2A). In an attempt to identify indirectly the cell type responsible for the synthesis of SGG, the labelling in the autoradiograms was quantitated and correlated with the stage in the cycle of the seminiferous epithelium (Fig. 3). Although incorporation of [35S]sulphate occurs in all tubule sections, there is, however, a quantitative difference in the labelling according to the stage of the seminiferous cycle. This is perhaps most obvious in the comparison of stage VII with stage XIII (Fig. 2B). Grain counting provides a more precise measure of this difference (Fig. 3). It must be emphasized that the sections have been categorized following staining with haematoxylin and eosin. Periodic acid-Schiff (PAS) staining for carbohydrate is the stain of choice for the morphological definition of the stages of the seminiferous epithelium (Parvinen, 1982). Unfortunately, this procedure oxidizes the silver grains and autoradiography is no longer possible. Moreover, the morphology of the tissue is not preserved in frozen sections to the degree necessary for unambiguous assignment of stages. It is therefore possible that the stage assigned to each section may in fact be one stage earlier or later. The morphology is consistent with the stage ascribed. In some cases (e.g. stages VII, VIII), an unambiguous assignment can be made. No sections were found that could be ascribed with any degree of certainty to stages I, IX, XI and XIV. SGG is the only testicular glycolipid labelled following in vivo administration of [ S]sulphate (Lingwood et al. 1981). The labelling within the tubule cannot be ascribed to protein labelling since it is lost after dehydration and fixation (Fig. 4). It is possible that sulphated glycosphingolipids are synthesized late in spermatogenesis (Lingwood, unpublished) but these are not detected by metabolic labelling (Lingwood et al. 1981). We therefore conclude that all the labelling within the tubule is due to synthesis of SGG. The labelling in the interstitial cells is not lost after dehydration and fixation (Fig. 4) and is therefore most probably due to the synthesis of sulphated glycosaminoglycans in these cells. The results shown in Fig. 3 suggest that the maximum synthesis of SGG occurs at stages XII—II. Previous studies have shown that the synthesis of SGG is not detectable in the rat until between 15 and 18 days after birth, corresponding with the appearance of primary spermatocytes, and that the level of SGG is constant from day 25 (Kornblatt et al. 1974). This indicates that the basal cells (spermatogonia) are not responsible for SGG biosynthesis. Moreover, neither later spermatocytes nor early spermatids are major sites of SGG synthesis (Letts et al. 1978). Thus, the increased synthesis of SGG at stage XII is most probably due to the appearance of zygotene

Testicular sulphogalactolipid biosynthesis

337

spermatocytes at this stage of the seminiferous epithelium. The high level of synthesis maintained until stage II indicates that the early pachytene spermatocyte stage is also a major site of SGG biosynthesis. This same conclusion was reached by Kornblatt (1979) from a kinetic analysis of the testicular incorporation of [35S]sulphate and the appearance of radiolabelled SGG in the epididymis. However, Kornblatt was not able to identify the cell type responsible for the onset of SGG biosynthesis. The present results indicate that stage X does not show increased SGG synthesis and, thus, SGG biosynthesis is not initiated at the leptotene spermatocyte stage. The present autoradiographic data support the contention that SGG is maximally synthesized in the zygotene and early pachytene spermatocytes and is within the boundaries of previous, less-precise estimations (Kornblatt et al. 1974; Kornblatt, 1979). The in vitro activity of the galactolipid sulphotransferase to make SGG correlates to some degree with the results of the autoradiographic study. Thus, the maximum specific enzyme activity is found in tubule sections containing stages IX-I (Table 1) when the in vivo labelling is highest, and in vitro enzyme activity is lowest when in vivo labelling is low (stages II—VI). In fact, this value approaches the lower limits of sensitivity of the enzyme assay under these conditions. However, considerable sulphotransferase activity was found for tubule segments containing stages VII and VIII when in vivo labelling was also low. These data suggest that during spermatogenesis, the galactolipid sulphotransferase is present before any SGG is synthesized. A similar conclusion can be drawn from the data of Kornblatt et al. (1974), who monitored the activity of the sulphotransferase and the level of SGG during the first wave of spermatogenesis. A developmentally regulated inhibitor of the testicular galactolipid sulphotransferase has been described and characterized. The inhibitor is first detected about 25 days after birth, corresponding to the appearance of mid-pachytene spermatocytes in the testis (Lingwood, unpublished). The tubule segments were assayed for the presence of inhibitor (Table 1). Maximum inhibitory activity was found at stages II—VI, which may be responsible in part for the minimal in vitro sulphotransferase activity detected at these stages when in vivo labelling, at least for stage II, remains high (Fig. 3). In summary, the present results have shown that the maximum synthesis of SGG occurs during a brief stage in spermatogenesis corresponding to the zygotene and early spermatocyte stages. Sulphotransferase activity in vitro is detected before the synthesis of SGG and may be terminated in vivo by the appearance of a sulphotransferase inhibitor at the mid-pachytene spermatocyte stage. I thank Dr I. Fritz, Best Institute, University of Toronto, for helpful discussions and assistance in learning the microdissection technique. This work was supported by MRC scholarship no. 9319, MRC grant no. MA7714, and NIH grant no. 2-P50HD 12629-04A1.

REFERENCES BRADFORD , M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analyt. Biochem. 72, 248-254.

338

C. A. Lingivood

KORNBLATT, M. J. (1979). Synthesis and turnover of e sulfogalactoglycerolipid, a membrane lipid, during spermatogenesis. Can.jf. Biochem. 57, 255-258. KORNBLATT, M. J., KNAPP, A., LEVINE, M., SCHACHTER, H. & MURRAY, R. K. (1974). Studies on

the structure and formation during spermatogenesis of the sulfoglycerogalactolipid of rat testis. Can. J. Biochem. 52, 689-697. LETTS, P. J., HUNT, R. C ,

SHIRLEY, M. A., PINTERIC, L. & SCHACHTER, H. (1978). Late

spermatocytes from immature rats. Isolation, electron microscopy, lectin agglutinability and capacity for glycoprotein and sulfogalactoglycerolipid biosynthesis. Biochim. biophys. Ada 541, 59-75. LlNGWOOD, C. A., HAY, G. & SCHACHTER, H. (1981). Tissue distribution of sulfolipids in the rat. Restricted location of sulfatoxygalactosylacylalkyl glycerol. Can.J. Biochem. 59, 556-563. LINGWOOD, C. A., SAKAC, D. & VELLA, G. J. (1983). Desulfation of sulfoglycolipids by anchimeric

assisted solvolysis. Carbohydrate Res. 122, 1-9. MURRAY, R. K., NARASIMHAN, R., LEVINE, M., SHIRLEY, M., LINGWOOD, C. A. & SCHACHTER, H.

(1980). Galactoglycerolipids of mammalian testis, spermatozoa and nervous tissues. ACS Symp., series no. 128, Cell Surface Glycolipids (ed. C. Sweeley), pp. 105-125. Washington: American Chemical Society Press. PARVINEN, M. (1982). Regulation of the seminiferous epithelium. Endocrine Rev. 3, 404-417.

{Received2 October 1984 -Accepted3 December 1984)