The deleterious action of doxorubicin on male germ cells has been described . The citotoxity caused by doxorubicin on the seminiferous epithelium can be related to its therapeutic activity; it interferes with molecules associated to the nuclear DNA and with enzymes (RNA and DNA polimerases, topoisomerases I and II) that are active in the cell division process. Then, doxorubicin forms a complex with chromatin , blocking the G2 phase of the cell cycle [47, 48] and provoking single and/or double strand DNA breaks . Doxorubicin also interferes with membrane lipids [50–53], provoking alterations in their chemical structure and impairing their function. The production of reactive species of oxygen, as a consequence of free radicals caused by doxorubicin, can also affect the cellular functions, altering the cellular metabolism in different aspects [52, 54]. In fact, the production of free radicals is one of the factors that limit the therapy with doxorubicin. This anticancer agent produces, as previously mentioned, a significant increase of lipid peroxidation and alterations of antioxidant enzyme activities in different organs of rats, including testis, as observed in vitro . It provokes significant elevation in the testicular malondialdehyde concentrations and decreases of glutathione content, glutathione reductase (GR), glutathione-S-transferase (GST), superoxide dismutase (SOD) and catalase activities, thus indicating oxidative stress production in doxorubicin-induced testicular toxicity . Besides, the deleterious effect of this anticancer agent on adult rat testis lipids and fatty acids has been observed after single or multiple dose regimes, resulting in a gradual loss of spermatogenesis and in a decrease in phospholipids, including glycerophospholipids and sphingomyelin; in this context, glycerophospholipids selectively loose their major polyunsaturated fatty acid (PUFA), while sphingomyelin looses its major very long-chain PUFA (VLCPUFA). By contrast, triglycerides and especially cholesterol esters (CE) tend to accumulate in the testes undergoing germ cell death, probably in the surviving Sertoli cells. Their fatty acid patterns suggest that initially these lipids retained part of the PUFA coming from, or no longer used for the synthesis of germ cell glycerophospholipids. Determining whether this accumulation results from a physiologic adaptation to the effects of doxorubicin or simply reflects another lipid derangement caused by the drug remains to be investigated .
Although some of the doxorubicin mechanisms of action on germ cell are known, the effects of this drug on other testicular components are poorly understood. Spermatogonia are the preferential doxorubicin target due to the presence of the enzyme topoisomerase II, but probably primary spermatocytes can also be damaged, although the role of topoisomerase II in pre-mitotic DNA synthesis, at determined stages, is more accentuated than in pre-meiotic synthesis as observed after etoposide-treatment . Thus, DNA synthesis in pre-meiotic spermatocytes is not so vulnerable to the doxorubicin action as pre-mitotic DNA synthesis. Moreover, it is possible that other topoisomerases can be involved in the process of pre-meiotic synthesis.
Thus, considering the different mechanisms of action of doxorubicin previously mentioned, it is also possible that additionally to spermatogonia, other cells, including Sertoli cells, are also targeted by doxorubicin.
In the present study, it was observed that doxorubicin administration to rats at early prepubertal phase alters transferrin production by Sertoli cells at specific phases of testicular development, indicating a functional alteration of these cells. On the other hand, transferrin synthesis by Sertoli cells is dependent on the presence of germ cells [28, 30, 57]. Thus, it is possible that the diminution of transferrin production by Sertoli cells is a consequence of germ cell depletion caused by doxorubicin. The significant recovery of transferrin synthesis by Sertoli cells and of the seminiferous epithelium observed in D127 subgroups support this hypothesis.
It is important to highlight that doxorubicin reduces the synthesis of the transferrin receptor and leads to atypical changes in intracellular iron distribution and trafficking . In the testes, transferrin receptors are present in the basal region of the Sertoli cell membrane and are essential to promote iron transportation from blood to the germ cells that are localized in the adluminal compartment of the seminiferous epithelium. Iron is crucial for germ cell proliferation and differentiation and Sertoli cells are the only way by which this ion can reach the germ cells at the adluminal compartment. To deliver iron ions to these cells, diferric plasma transferrin is endocytosed through the receptor at the basal region of the Sertoli cell membrane. In the cytoplasm of the Sertoli cells, iron is detached from plasma transferrin, captured by Sertoli cell transferrin and delivered to germ cells . Thus, if transferrin receptors are damaged by doxorubicin, the iron traffic through the seminiferous epithelium could have been affected. In addition, the germ cells localized in the basal compartment of the seminiferous epithelium get iron directly from plasma transferrin . Thus, damages to transferrin receptors could also cause reduction of iron capture by these cells. Indeed, in the D40 subgroup, spermatogonia were negative for transferrin, what may have contributed to the reduction of the volume density of transferrin-positive seminiferous epithelium.
In addition to transferrin receptor alterations, doxorubicin also increases the plasmatic levels of transferrin . Since most of the transferrin present in the interstitial tissue comes from the blood stream, it could be possible that the increase in the volume density of transferrin-labeled interstitial tissue observed in the D64 subgroup was a result of an increase of plasmatic transferrin levels. However, it does not seem to be the case, since the total volume of the interstitial tissue also increased. Moreover, the volume density of transferrin-positive interstitial tissue of D40 and D127 subgroups was normal, indicating that transferrin of this testicular compartment was not affected. This suggests that, in the testis, doxorubicin acted specifically on Sertoli cell transferrin.
Although the damages to Sertoli cell function are generally reversible, the morphological alterations observed after doxorubicin treatment suggest that these cells may have been directly injured in addition to their secondary damage occurred due to the germ cell primary harm. The presence of Sertoli cell nuclei in the tubular lumen, for example, indicates that the structural integrity of these cells was affected by the treatment with doxorubicin. Dislocation of Sertoli cell nuclei from basal to adluminal or luminal compartments has also been demonstrated in adult rats treated with cimetidine [62, 63]. In these studies, this alteration was associated to Sertoli cell death. Morevover, in another research, reduction of Sertoli cell number followed by a decrease in sperm production, normal morphology and motility was observed in doxorubicin-treated adult mice, 6 weeks after the end of the treatment . In the present study, although we did not score the Sertoli cell number, the Sertoli cell nuclei detached into the lumen indicate that these cells may have been drastically injured. Thus, besides the secondary Sertoli cell damage, due to the primarily occurred germ cell death, it is also important to consider that when doxorubicin was administered, Sertoli cells were passing through critical phases of their development, what makes a primary damage more likely to occur. Around 15dpp, when the rats received the first dose of doxorubicin, Sertoli cells stop to proliferate  and the blood-testis barrier start to be formed [66, 67]. At 22dpp, when the second dose was administered, Sertoli cells were still undergoing maturation [65, 68]. Therefore, we could also consider the higher susceptibility of Sertoli to doxorubicin at these phases than adult Sertoli cells. Because Sertoli cells are crucial for spermatogenesis, damages in these cells at early pubertal phase could also lead to germ cell death in other later periods of sexual maturation as peripuberty (40 days) and after the completion of puberty (64 days). Hence, considering the Sertoli cell morphological alterations observed in this study, it is possible that the seminiferous epithelium alterations observed in the doxorubicin-treated rats could also be consequence of direct Sertoli cell damage. Indeed, some alterations such as intraepithelial vacuolization, spermatid retention and high frequency of Sertoli cell nuclei in which the nucleolus was not evident suggest that Sertoli cells were damaged independently of germ cell death. Moreover, despite the possibility of the decrease of Sertoli cell transferrin labeling a consequence of germ cell depletion, it is important to consider that doxorubicin can increase the production of free radicals as previously observed [52, 54]. In addition, primary immature Sertoli cell obtained from 18-day-old rat testes and cultured with the anticancer agents cis-diamminedichloroplatinum (CDDP), adriamycin and vinblastin revealed that these agents have direct damaging effects on rat Sertoli cell, decreasing the level of transferrin. In this research, the concentration of transferrin in the culture medium was measured and used as an indicative of Sertoli cell function . The role of Sertoli cell in postchemotherapy azoospermia has also been noticed in a 31-year-old patient who underwent cancer cytotoxic chemotherapy for non-Hodgkin's lymphoma at 13 years of age . In this patient, a fraction of Sertoli cells (13%) in the atrophic tubules re-expressed the intermediate CK-18 filament protein, which is normally absent after puberty, but not the D2-40 antigen, a membrane-linked glycoprotein which loss of expression at puberty marks an irreversible step in Sertoli cell maturation. The reversion to a dedifferentiated state, marked by the reexpression of CK-18 as a consequence of chemotherapy, besides the partial inactivation of Sertoli cells following the chemotherapeutic drug cytotoxicity may contribute to the spermatogenic impairment, then resulting in infertility . Although testicular germ cell products can regulate Sertoli cell function [71, 72] and alter the production of transferrin, for example, it is also possible that a harmful effect of doxorubicin on Sertoli cell might have occurred in the present study. Disruption of Sertoli cell structure and shedding of immature germ cells have been observed in doxorubicin-treated adult mice . However, other experiments using labeling of house-keeping proteins such as actin and/or markers of Sertoli cell differentiation as cytokeratin-18 must be conducted to better clarify this subject. Another relevant hypothesis is that the blood-testis barrier injury, caused by doxorubicin toxicity, was mediated by the generation of free radicals [1, 2] and lipid peroxidation . In fact, studies in the testis and other organs have illustrated the role of environmental toxicant-induced oxidative stress in mediating the disruption of cell junctions, which is regulated by the activation of phosphatidylinositol 3-kinase (PI3K)/c-Src/focal adhesion kinase (FAK) and mitogen-activated protein kinase (MAPK), signaling pathways involving polarity proteins and leading to reproductive dysfunction, such as reduced sperm count and semen quality in men . However, the impact of doxorubicin toxicity on integrity and damage of the blood-testis barrier during prepuberty are still to be established.
Important alterations were observed in the frequency of the seminiferous epithelium cycle after doxorubicin treatment, especially at 64dpp. The seminiferous epithelium cycle is a strictly controlled process that is characterized by specific germ cell associations, defined as stages of the seminiferous epithelium cycle. During this cycle, Sertoli cells change their morphology and function, according to the requirements of the spermatogenic process. Because Sertoli cells are responsible for the synchronization of the seminiferous epithelium cycle, alterations of these cells can cause problems to the progression of the stages during the cycle. It is also important to consider that the massive loss of germ cells disturbs the typical cell association of each stage, leading to alterations of the frequency of the stages of the seminiferous epithelium cycle. Another important factor that should be considered is that postpubertal and adult doxorubicin-treated rats showed retention of step 19 spermatids. At this step, these cells are released into the tubular lumen through a process called spermiation, which occurs at stage VIII of the seminiferous epithelium cycle. This process is controlled by Sertoli cells  and injuries to these cells can alter spermiation and cause spermatid retention .
Alterations of the seminiferous epithelium cycle has been described after administration of chemicals such as 1, 3 dinitrobenzene  and 2, 5 hexanedione , which are referred as Sertoli cell toxicants [77, 78]. In general, chemotherapeutic drugs are not considered Sertoli cell toxicants. However, previous studies by our group have suggested that etoposide, another chemotherapeutic drug, in addition to causing damage to the germ cells, may also provoke direct damages to Sertoli cells [17, 18]. The present study also points to a possible effect of doxorubicin on Sertoli cells. Indeed, some Sertoli cell alterations suggest that these cell damages are more severe than those considered be exclusively secondary effects resulted from germ cell death.
Another important aspect is that the stage-specific gene expression is a fundamental characteristic of rat spermatogenesis and Sertoli cells . In fact, in adult doxorubicin-treated mice, a quantitative RT-PCR analysis showed a dysregulation in the expression of some genes such as Csk and Axl, which are important to the remodeling of seminiferous tubule during spermatogenesis and to the germ cell differentiation respectively . These remarks could support some of our observations, concerning the conspicuous alterations in the frequency of some seminiferous epithelium stages in doxorubicin-treated rats, in all ages investigated (40, 64 and 127 days). In addition, the inactivation and delay of the Sertoli cell maturation due to cytotoxicity of the chemotherapeutic drugs may contribute to the spermatogenic impairment  and could be related to the functional alterations of the Sertoli cell, and probably to the changes in seminiferous epithelium cycle as well. In our report, the delay or interruption of the Sertoli cell differentiation could justify, at least in part, the significant increase of the frequency of stage I and the reduction of subsequent stages such as II -III, IV, V, VI at 40 days, in comparison to the control rats. In addition, it has been shown that the aforementioned stages II and III are infrequently pinpointed as being especially vulnerable to agents that act on spermatogenesis ; therefore, it is possible that their frequencies have been altered due to doxorubicin direct action on Sertoli cells. Moreover, as previously mentioned, it is important to remember that doxorubicin was administered during prepubertal phase, when the Sertoli cells were still undergoing maturation.
Summarizing, doxorubicin is a very potent drug that acts through different mechanisms of action. Without doubt, a secondary damage of Sertoli cell occurred due to the injury caused to the germ cells. On the other hand, the alterations observed in the present study, along with the fact that Sertoli cells were not completely mature when doxorubicin was administered, suggest that the direct damage to the Sertoli cell observed is likely to be also responsible, at least in part, for some of the testicular alterations noticed. The iron atypical chelator action of doxorubicin, which provokes the decrease of transferrin receptor synthesis, leading to atypical changes in intracellular iron distribution and trafficking , can also alter the synthesis of trasferrin by the Sertoli cell, a phenomenon that should also be considered.
Measurements of transferrin contents in rat testes can indicate damage to Sertoli cell function. High doses of cisplatin (8 mg/kg), for example, affect testicular transferrin concentration, but lower doses (4 mg/kg and 2 mg/kg) have no significant effect on Sertoli cell function. Thus, an anti-cancer agent primarily may affect the DNA synthesizing activity of spermatogonia and spermatocytes, but high doses of these agents have deleterious effects on Sertoli cells . The age of treatment chosen can also be a determining factor in the type of testicular damage observed. However, detailed studies will be necessary to verify the direct damage of Sertoli cell by the anticancer agent doxorubicin when administered in early prepubertal rats.