Impairment of rat spermatogenesis following unilateral experimental ischemia

Transcription

1 FERTILITY AND STERILITY Copyright 1982 The American Fertility Society Vol. 38, No.4, October 1982 Printed in U.8A. Impairment of rat spermatogenesis following unilateral experimental ischemia David Iwan Lewis-Jones, M.B., Ch.B. Maria Moreno de Marval, M.Sc. * Ronald George Harrison, M.A., D.M.t Department of Anatomy, University of Liverpool, Liverpool, England When one testis in a rat is rendered ischemic, histologic damage occurs in the contralateral testis, which can be explained on immunologic grounds. An increase in the quantity of a naturally circulating cytotoxic antitestis antibody or an alteration in the specificity of this antibody was observed via cytotoxic antibody tests. Immunocytochemical techniques further demonstrated this phenomenon, showing binding of fluorescein-labeled antibody to sections of normal control testis or to sections of contralateral testis. It appeared that the ischemic testis presented altered ischemic cells or normal cells to the immune system as its blood-testis barrier was broken down. It is suggested that to leave a necrotic testis within the scrotum following prolonged torsion of the testis in the human may be a reason for subfertility or infertility in this type of condition, and a similar result may occur following unilateral testicular damage caused by varicocele or orchitis. Fertil SteriI38:482, 1982 Infertility due to autoimmunity to spermatogenesis has been shown to occur in animals under experimental conditions, and autoimmunity can also affect seminiferous tubule architecture and function. 1-9 Previous work by Harrison et al. 10 demonstrated that damage to testicular structure can occur as a result of the injection of ischemic testicular tissue as a homogenate or the grafting of ischemic testis into rats. When torsion of the testis was mimicked in rats by the ligation of the testicular artery and vein in the abdomen between ligatures, not only was this Received January 27, 1982; revised and accepted June 8, *Present address: Departmento de Morfologia, Escuela de Medicina, Universidad de Oriente, Venezuela, South America. treprint requests: Ronald George Harrison, M.A., D.M., Department of Anatomy, University of Liverpool, P.O. Box 147, Liverpool, L69 3BX, England. ipsilateral testis rendered ischemic to the extent that all spermatogenic tissue was destroyed within 1 week, but the contralateral testis was also considerably damaged. In common with other workers, we observed degenerative changes not only in the germinal cells of the contralateral testis but also in the Sertoli cells and in the membrana propria of the seminiferous tubules. 1l - 14 It became apparent that the possible immunologic basis of these findings required analysis, because the serum from these animals was cytotoxic to normal rat spermatozoa and sections of normal rat seminiferous tubules and sections of contralateral testes from the test animals showed uptake of antibody by the direct immunofluorescence method. MATERIALS AND MEmODS Littermate-inbred, sexually mature Wistar albino rats varying in weight from 200 to 450 gm 482 Lewis.Jones et ai. Impairment of spermatogenesis by ischemia Fertility and Sterility

2 ~ " After 7 days all animals were sacrificed by an overdose of ether, and their testes were removed and placed in Bouin's fixative. Small pieces of these testes were also placed in alcohol, formal calcium, or Karnovsky's fixative. 15 Blood was removed from each animal by cardiac puncture, placed in a nonheparinized tube, and the serum was removed 2 hours later by pipetting off the supernatant from the cells after centrifugation at 100 x g for 5 minutes. These serum samples were then frozen at - 20 C until required. Figure 1 Spermatogenic destruction seen in a seminiferous tubule of a ligated testis showing destruction of spermatogonia and Sertoli cells and separation of the membrana propria (m) from spermatogenic cells (original magnification, x 100). were chosen for this experiment and divided into three groups: (A) ischemic group (ten rats), (B) operation control group (ten rats), and (C) a group of rats as a source of normal rat serum and normal testis as an additional control (ten rats). The group A rats were anesthetized by ether, and the abdominal cavity was opened by a low central midline incision. The right testis was removed from the scrotum, and the vascular bundle was isolated and divided between two silk 5/0 sutures just above its anastomosis with the deferential artery. The testis was then returned to the scrotum. The vascular supply of the left testis was examined for abnormalities, and care was taken to cause no mechanical damage to it. Any animals found with an anomalous blood supply were removed from the experiment. The group B rats were subjected to the same exploratory operation, and the vascular pattern to both testes was examined for abnormalities. No experimental ischemia was performed in this group, and they acted as an operation control group. HISTOLOGIC INVESTIGATION The testes fixed in Bouin's solution and embedded in paraffin were cut at 8 ~m (1 in 10) and stained with Weigert's iron hematoxylin and eosin and Weigert's iron hematoxylin and van Gieson's stain (for collagen fibers). In order to examine the nature of histologic alterations, especially in the membrana propria of the seminiferous tubule and of cells in the interstitial tissue, pieces of testis were fixed in alcohol arid after embedding in paraffin were stained with periodic acid-schiff and methyl green pyronine. To demonstrate the presence of lipids, testicular tissue was fixed in Baker's formal calcium, embedded in gelatin, sectioned at 10 ~m on a freezing microtome, and colored with Sudan red and Mayer's hematoxylin or Sudan black B. Figure 2 Thickened tunica albuginea (T) seen in the ischemic testes with desquamation of its mesothelium and a surrounding infiltration of round cells and fibroblasts (original magnification, x 100). Vol. 38, No.4, October 1982 Lewis-Jones et ai. Impairment of spermatogenesis by ischemia 483

3 (LKB, Produkter AB, Bromma, Sweden), doublestained with lead and uranyl acetate, and studied with an electron microscope (Philips 300, Eindhoven, The Netherlands). IMMUNOLOGIC INVESTIGATION Two types of immunologic tests were performed, the cytotoxic test and the indirect immunofluorescence technique. Cytotoxic Antibody Heat-inactivated serum from animals in all groups was serially diluted in two volumes of diluent (45% Medium 199, 45% Hanks' solution, 5% inactivated fetal calf protein, and 5% phosphate buffer, ph 7.2). To each dilution of serum was added one volume of complement (fresh guinea pig serum diluted 1 in 5) and one volume of normal rat spermatozoa obtained from the epididymis of rats from the same colony.7 The dilutions were incubated at 37 C for 30 ~inutes and then scored for percentage of dead spermatozoa evalu-. ated by eosin and trypan blue penetration 16 and a count made of the motile spermatozoa. Results were obtained by two observers working simultaneously through the same microscope. Controls against serum without complement, complement only, and diluent only were used. Figure 3 Histologic section of a contralateral testis with varying degrees of impairment to spermatogenesis (original magnification, x 40). ELECTRON MICROSCOPY Small pieces of tissue were fixed for 3 hours at 4 C in a mixture of 4% glutaraldehyde and 2% paraformaldehyde in phosphate buffer (modified from Karnovsky15) and adjusted to ph 7.4 with hydrochloric acid. Mter three to five changes of phosphate buffer, the pieces were postfixed in 1% osmium tetroxide in phosphate buffer. Following dehydration, the pieces of testis were placed in propylene oxide for three periods of 10 minutes, followed by propylene oxide and resin-25% resin for 1 hour, 50% resin overnight, 75% resin for 1 hour-and then transferred into fresh resin and placed in a 60 C oven for 48 hours. Sections 50 to 90,...m thick were cut on an ultramicrotome Figure 4 A seminiferous tubule stained with periodic acid-schiff showing a thickening of the membrana propria (m) and vacuolation between germinal cells. The maturation of spermatogenic cells appears grossly damaged (original magnification, x 100). 484 Lewis-Jones et al. Impairment of spermatogenesis by ischemia Fertility and Sterility

4 the controls. Natural background fluorescence seen in normal testicular tissue could not be eliminated, because it was seen in sections treated only with saline or buffer, and this had to be taken into account when we were assessing the test sections. RESULTS HISTOLOGIC RESULTS The histologic picture of normal rat testis is well documented,19, 20 and these findings were verified by examination of normal control testes from the rats sacrificed for their serum. In group A both the ipsilateral ischemic testes and the contralateral testes were examined. Ipsilateral Ischemic Testes (Group AJ. The testes from all ten rats were shrunken, and their histologic appearance was in accordance with the findings of OettIe and Harrison. ll Severe spermatogenic damage was seen in tubules near the periphery of the testes, but in most other tubules Figure 5 Electron micrograph of a contralateral seminiferous tubule having a thickened irregular membrana propria (m) with an increase in collagen in the internal lamella (C). The Sertoli cell nucleus (8) is abnormal, its cytoplasm is vacuolated, and its mitochondria are abnormal (original magnification, x 4320). I mmunof/uorescence For the indirect immunofluorescence technique, the sera from all three groups were diluted 1 in 8 with phosphate-buffered saline and applied to dried sections of normal testicular tissue, sections of paraffin-embedded normal testicular tissue, dried sections of contralateral testis from the animal whose serum was being tested, and paraffin-embedded sections of the contralateral testis17 in a covered and moistened container for 40 minutes.18 Sections were washed by agitation in buffer and incubated with fluorescein-labeled antirat gamma globulin dilutedl in 60 for 30 minutes. After rinsing, they were mounted in glycerol. Sections were examined with a fluorescent microscope (Vickers Ml7, Manchester, England) and scored for their degree of fluorescence from 20 areas picked randomly from each section. The exposure time given on the microscope was for the whole field of view, and this value (in seconds) gave a semiquantitative estimation of the degree of brightness in test sections when compared with Figure 6 Electron micrograph of a contralateral seminiferous tubule showing a disorganized spermatocyte (8p) with abnormal clumping of the chromatin granules in the nucleus and a reduction in the amount of cytoplasm, which also contains abnormal vacuolated mitochondria. Also present is a Sertoli cell nucleus (8) nearer the lumen than usual. Numerous large lipid droplets are also seen (L) (original magnification, x 3926). Vol. 38, No.4, October 1982 Lewis-Jones et ai. Impairment of spermatogenesis by ischemia 485

5 Figure 7 Electron micrograph of an early spermatid (Sm) showing karyolysis and disorganization of the nuclear chromatin. In the cytoplasm mitochondria are degenerated and vacuolated. There is vacuolation (V) between the cytoplasmic membrane and the adjacent Sertoli cell cytoplasm (original magnification, x 4320). spermatogenic destruction had occurred (Fig. 1). In the latter case all spermatogenic cells were destroyed, even the spermatogonia and Sertoli cells, some of which had become macrophages. The lumen in some tubules was filled by the cytoplasm and nuclei of dead cells. In other tubules germinal cells were still present, but they were phytoid, hyperchromatic, and shrunken. They had often become separated from the membrana propria, and exfoliation into the lumen had occurred. The primary spermatocytes were hypervesicular, the spermatids had a ragged cytoplasm, and some multinucleate spermatids were present. In some tubules multinucleate primary spermatocytes were present. Most cells remaining showed pyknosis and karyolysis with densely clumped chromatin in resting spermatogonia. As a result of exfoliation, some cells such as early spermatids and primary spermatocytes could be seen in the lumen. Many tubules were filled with necrotic cells which had rounded off and lost their normal nuclear staining. Hyperchromatic hypertrophied Leydig cells were seen in the interstitial spaces, and in some areas there was an infiltration of round cells and fibroblasts. The tunica albuginea appeared thickened, and there was desquamation of its mesothelium (Fig. 2). Contralateral Testes (Group A). A varying degree of impairment to spermatogenesis was observed in all ten testes. In 30% of tubules very little damage was observed, but in the majority (70%) there was pronounced spermatogenic damage (Fig. 3). In the damaged sections the membrana propria appeared thickened and to have become separated from underlying cells in certain areas; vacuolation was often seen in this region within the basal Sertoli cell cytoplasm (Fig. 4). There was a diminished number of spermatozoa in some tubules, and pyknotic bodies were frequently seen near the lumen, which stained with Sudan red or black. In the tubules showing greatest damage, the early spermatids were abnormally elongated and necrobiotic, and the more mature spermatids, when present, were pyknotic and distorted. Exfoliation of cells into the lumen was a common finding, so that cells at different stages of spermatogenesis were seen filling the lumen, mixed with pyknotic bodies (Fig. 3). In the more damaged tubules necrotic cells that had rounded off and lost their normal nuclear staining were seen. In the interstitial spaces the Leydig cells appeared normochromic but hypertrophied; although there was no inflammatory reaction present, there was an increase in cells with the characteristics of plasma cells. The tunica albuginea appeared normal. Figure 8 Electron micrograph of a Leydig cell (L) in a contralateral testis showing a hypertrophic nucleus with abnormal clumping of its chromatin (original magnification, x 4320). 486 Lewis-Jones et al. Impairment of spermatogenesis by ischemia Fertility and Sterility

6 Table 1. The Relative Cytotoxicity of Serial Dilutions of the Test Sera Against Normal Rat Spermatozoa as Indicated by Eosin and Trypan Blue Penetration Mean percentage of spermatozoa dead after exposure to the serum dilution Test serum /192 1/96 1/48 1/24 Normal serum (ten samples) Operation control (five samples) Ligated serum (ten samples) acontrol inactivated serum without addition of complement. b Addition of complement only. CAddition of diluent only. Control Control Control with se- withco~le- with dil- 1/ rum only" mento yb uent only" Operation Control Testes (Group B). The histologic picture in this group appeared fairly normal in the five animals examined. Ten percent of the tubules showed a small amount of damage, but never as severe as in the ipsilateral or contralateral testes of group A. ELECTRON MICROSCOPY The ultrastructure of both normal and ischemic testes is well documented. 14 All seminiferous tubules in the contralateral testes of group A showed a thickened membrana propria with an increase of collagen fibers in the internal lamella (Fig. 5) and an increase in pinocytic activity in the internal and external cellular layvrs. The membranes also appeared more irregular than normal. The Sertoli cell nuclei had a variation in the chromatin distribution that was more granular than normal. Most nuclei displayed areas of loss of nuclear material. The Sertoli cell cytoplasm showed a generalized increase in vacuolation and lipid formation. There was an abundance of dense bodies (lysosomes). The mitochondria showed an increase in pleomorphism, and many showed vacuolation. In some sections Sertoli cells were found rather randomly distributed and in general nearer the lumen of the seminiferous tubules than usual (Fig. 6). Lamellar dense bodies were usually found in the luminal cytoplasm of the Sertoli cells. The spermatogonia were in general retracted, with loss of cytoplasm, and the remaining cytoplasm showed small vacuoles (normally none). The spermatocytes displayed karyolysis, greatly reduced cytoplasm, and a decrease in the number of mitochondria, some of which were also vacuolated; a small amount of vacuolation was seen in the cytoplasm itself. The early and late spermatids were abnormally shaped. The nuclear membrane was disrupted in places. In the cytoplasm the mitochondria were disorganized in structure, with severe vacuolation and degeneration, and their distribution within the cytoplasm was irregular. All of the more mature spermatids had a degeneration of the nucleus and acrosome, with loss of nuclear and cytoplasmic material (Fig. 7). Leydig cells showed increased nuclear size and chromatin in clumps, but the cytoplasm appeared relatively normal (Fig. 8). There was also a moderate increase in collagen fibers in the interstitial tissue. CYTOTOXIC TEST RESULTS The results of testing the antisera for their relative cytotoxicity and immobilizing effect are shown in Tables 1 and 2 and in Figure 9. Normal rat serum is to some extent cytotoxic to spermatozoa and reduces their mobility, presumably protected in vitro by the "blood-testis barrier." In the operation control group (group B) the serum was more cytotoxic as a result of the manipulation of the vascular supply to the testes alone. In group A (unilateral testicular ischemia) the serum was found to be the most cytotoxic and to have a greater immobilizing effect. In controls using diluent only and antiserum but not complement, the spermatozoa stayed alive longer than 4 hours, with very few dead cells and with a higher proportion of motile cells. IMMUNOFLUORESCENCE RESULTS The rat testis has a degree of "natural" or "background" fluorescence that has to be taken into account while interpreting degrees offluorescence in test sections. In the control groups (both saline and phosphate buffer) there was no concen- Vol. 38, No.4, October 1982 Lewis-Jones et al. Impairment of spermatogenesis by ischemia 487

7 Table 2. The Relative Immobilizing Effect of Treating Normal Rat Spermatozoa with Serial Dilutions of the Test Sera Mean percentage of spermatozoa motile or mobile after exposure to the serum diluted Test serum / / Normal serum (ten samples) Operation control (five samples) Ligated serum (ten samples) acontrol inactivated serum without addition of complement. b Addition of complement only. CAddition of diluent only. Control Control Control with Be- withco~le- with dil 1/12 1/6 1/3 rum only" mento yb uent only" trated fluorescence in any cell and no clear staining of the membrana propria, and the appearance was of a faded outline of cells arranged in the normal manner. When sections of contralateral testes from group A, testes from group B, and normal rat testes were examined following the application of fluorescein-labeled anti-gamma globulin, the fluorescence results were similar to those of normal testis sections that had been treated with the test sera and then with the fluorescein-labeled antigamma globulin. Tissue fluorescence was brighter and better preserved in paraffin sections. Group A (ischemia) serum appeared to cause the greatest fluorescence. Group B (operation control) serum gave less fluorescence than group A but more than normal rat serum. The control sera gave the least fluorescence. Following absorption of the group A serum with normal spermatozoa, the fluorescence obtained became similar to that seen in the controls. In the test sections of group A the membrana propria and all spermatogenic celis showed considerable fluorescence. Which cells were stained in an individual seminiferous tubule depended on the stage of spermatogenesis observed. In those sections containing a greater number of mature cells, the staining appeared generally brighter. The spermatogonia and spermatocytes showed a yellow-green cytoplasm contrasting with the nonfluorescent nuclei. The mature spermatids were brighter than any other element (Fig. 10). In some cases the cytoplasm of membrana propria cells, spermatogonia, and spermatocytes exhibited fluorescent granule-like structures. In the interstitial spaces, cells appearing to be Leydig cells also showed a stained ring cytoplasm. Other cells with characteristics of plasma cells were similarly stained. DISCUSSION When ischemic testes are either grafted or injected as a homogenate into littermate rats, there is considerable damage to the testes of the host animals.lo Previous work has shown that iso- or autoimmunization of the guinea pig by testis or spermatozoa can lead to testicular damage. 7,8 Initially, it was assumed that the damage was mediated by immune cells, but it is now thought that both cellular and humoral responses are necessary.4,6 The damage caused by the humoral response is probably by the IgG and IgM antibody classes. 21 It seems that the injected or grafted testicular tissue, whether normal or ischemic, acts as an antigen and causes the immune response in the host. Many attempts have been as o N o a; E.. CD ~ II) 100 -goo CD "t:i '0 50 CD ~40 1: ~30 ~ ,, "-..., '.., " " " Figure 9 The percentage of dead control spermatozoa resulting from treatment with different dilutions of the test sera...., '.. "" Ligated '.serum (group A) ''a--- Operation "'II control (group B) ~ \ \, \2 ~4 ~8 ~\92k4 Serum dilution Normal serum 488 Lewis-Jones et a1. Impairment of spermatogenesis by ischemia Fertility and Sterility

8 Figure 10 Indirect immunofluorescence technique showing uptake of antibody from the test sera by the membrana propria (m), the spermatogenic cells (8), and Leydig cells (L) (original magnification, x 40). made to isolate and purify the spermatogenic antigens from germinal cells and from spermatozoa.6,22 When one testis in a rat is rendered partially ischemic, not only does this testis become necrotic but its contralateral testis also shows spermatogenic damage. In view of the similar findings (both histologically and immunologically) in this type of experiment, when compared with the testes of rats that have received either injections of ischemic testis homogenate or grafts of ischemic testis, it is postulated that the damage to the contralateral testis is also an immunologic phenomenon. Normally circulating in a rat's serum are immunoglobulins, some of which are naturally cytotoxic to the contents of the seminiferous tubules, spermatogenic cells, and spermatozoa, as shown in the cytotoxic and immunofluorescence experiments. Some barrier in or around the seminiferous tubule, such as the membrana propria, normally protects it from the effects of the serum in vivo. 9 When the antigen (in these cases an ischemic testis, either as the ipsilateral testis in the same animal or as a graft or homogenate injection from a littermate) is presented to the system, the "blood-testis barrier" in the contralateral testis breaks down and antibodies are able to enter the seminiferous tubule and damage both spermatogenic and nonspermatogenic cells. The nontubular interstitial cells are also damaged. The increase in the cytotoxicity of the serum as a response to operating on the animal and manipulating the blood supply may be due to minor testicular damage as a result of this procedure. The stress to the system of the operation alone is also sufficient to raise the level of naturally circulating antibodies. It is difficult to say at present which component of the ischemic testis acts as the antigen. A combination of all types of cells may, once rendered degenerate, release potentially antigenic material, resulting in antibody formation. Perhaps other cells in isolation such as Leydig cells, Bertoli cells, spermatogonia, or the cells forming the layers of the membrana propria may be responsible. The antigen itself which causes the immunologic response may not necessarily be degenerate. Normal or slightly damaged germinal cells or cells from the membrana propria may be exposed to the immunologic system as the ischemia causes breakdown of the blood-testis barrier. Which type of cell is affected first in the contralateral testis is difficult to assess. Damage to the membrana propria visible on electron-microscopic examination and in the indirect immunofluorescence test suggest that this is probably the defect allowing antibody to reach the spermatogenic cells. Because these latter cells are naturally antigenic, the antibody is therefore able to affect them. Damage at the rete testis is probably greatest, because it is here that the blood-testis barrier is weakest. 9 This would not, however, explain how or why the Leydig cells are affected. Histologically they appear hypertrophic, and they are also stained by the immunofluorescence technique. It is well documented that the Leydig cells, via their production of testosterone, influence spermatogenesis; it is therefore important to determine whether an experimental increase in immunoglobulins or an altered immunoglobulin may affect androgen levels. Whether changes in the Leydig cells are involved directly with the immunologic processes is unknown at present. The clinical aspects of this work relate to treatment of torsion of the testis in the human. A necrotic testis as a result of untreated unilateral torsion corresponds with the ligated ischemic testes in our experiments. If this twisted testis is left in situ, as is often the case when the testis is thought clinically nonviable, it would seem that Vol. 38, No.4, October 1982 Lewis-Jones et ai. Impairment of spermatogenesis by ischemia 489

9 it would cause damage to the contralateral unaffected testis. This may be a cause of subfertility or infertility following torsion of the testis. If testicular torsion is suspected, the testis must always be explored, however nonviable it is thought to be clinically. If found viable, it should be untwisted and then tethered, and a similar tethering operation should be performed on the contralateral side. If found nonviable, it should be removed, and a tethering procedure should be performed on the contralateral side. A twisted testis, however nonviable, should always be explored and removed once deemed degenerate. If it remains, it allows a continued antigenic stimulus to the immune system. Antibodies thus produced would continue to attack the contralateral testis until the antigenic stimulus reduces as the affected testis become fibrotic. Active intervention is always indicated in such cases. Infertility in men with unilateral testicular damage, for example, following varicocele or orchitis, is also now more understandable. In the case of varicocele, it is thought that the retrograde blood flow through the testicular vein causes damage to a testis as a result of disturbed thermoregulation. 23 The damage to the contralateral side is also thought to be due to this disturbed temperature phenomenon. This research suggests that the testis damaged by the effects of a varicocele will set up an autoimmune phenomenon that, possibly in combination with a change in temperature or alone, will damage the normal unaffected contralateral testis. In orchitis there is widespread damage to a testis if the condition is unilateral, and this condition may also initiate an autoimmune reaction that damages the unaffected testis. REFERENCES 1. Freund J, Lipton MM, Thompson GE: Aspermatogenesis in guinea-pig induced by testicular tissue and a<\iuvant. J Exp Med 97:711, Boughton B, Spector WG: Autoimmune testicular lesions induced by injury to the contralateral testis and intradermal injection of a<\iuvant. J Pathol 86:69, Bishop DW, Carlson GL: Immunologically induced aspermatogenesis in guinea-pigs. Ann NY Acad Sci 124:247, Brown PC, Holborow EJ, Glynn LE: The aspermatogenic antigen in experimental allergic orchitis in guinea-pigs. Immunology 9:255, Menge AC: Induced infertility in cattle by iso-immunisation with semen and testes. J Reprod Fertil 13:445, Voisin GA, Toullet F: Etude sur l'orchite aspermatogenetique autoimmune et les autoantigenes de spermatozoides chez Ie cobaye. Ann Inst Pasteur 114:727, Johnson MH: Changes in the blood testis barrier of the guinea pig in relation to histological damage following iso-immunization with testis. J Reprod Fertil 22:119, Johnson MH: Selective damage to spermatogenic cells of high antigenicity during auto-allergic aspermatogenesis. J Pathol 102:131, Johnson MH: An immunological barrier in the guinea pig testis. J Pathol 101:129, Harrison RG, Lewis.Jones DI, Moreno de Marval MJ, Connolly RC: Mechanism of damage to the contralateral testis in rats with an ischaemic testis. Lancet 2:723, Oettle AG, Harrison RG: The histological changes produced in the rat testis by temporary and permanent occlusion of the testicular artery. J Pathol 64:273, Steinberger E, 'fjioe DY: Spermatogenesis in rat testes after experimental ischemia. Fertil Steril 20:639, 'fjioe DY, Steinberger E: A quantitative study of the effect of ischemia on the germinal epithelium of rat testes. J Reprod Fertil 21:489, Kaya M, Harrison RG: An analysis of the effect of ischemia on testicular ultrastructure. J Pathol117:105, Karnovsky MJ: A fixative of high osmolarity for use in electron microscopy. J Cell BioI 27:137a, Boyse EA, Old LJ, Chouroulinkov I: Cytotoxic. test for demonstration of mouse antibody. In Methods in Medical Research, Vol 10, Edited by HN Elsen. Chicago, Year Book Medical Publishers Inc, 1964, p Sainte-Marie G: A paraffin embedding technique for studies employing immunofluorescence. J Histochem Cytochem 10:250, Nairn RC: Fluorescent protein staining Fourth edition. Edinburgh, Churchill Livingstone, 1976, p Leblond CP, Clermont Y: Spermatogenesis of rat, mouse, hamster and guinea pig as revealed by the "periodic acidfuchsin-sulfurous acid" technique. Am J Anat 90:167, Courot M, Hochereau-de Reviers MT, Ortavant R: Spermatogenesis. In The Testis, Development, Anatomy & Physiology, Edited by AD Johnson, WR Gomes, NL Vandemark. London, Academic Press, 1970, p Johnson MH: Characterization of a natural antibody in normal guinea pig serum reacting with homologous spermatozoa. J Reprod Fertil 16:503, Katsh S, Katsh G: Antigenicity of spermatozoa. Fertil Steril 12:522, Zorgniotti AW: Testis temperature, infertility and the varicocele paradox. Urology 16:7, Lewis-Jones et al. Impairment of spermatogenesis by ischemia Fertility and Sterility