3.1. Quality assessment of the identified studies
The quality of 52 selected studies were assessed using ToxRTool. Four records were analysed as poor studies because of missing material characterization and/or usage of ultra-high concentrations or doses and/or no controls for biological endpoints have been included. Seven records were acceptable studies, with most quality criteria fulfilled but not strictly all. Finally, forty-one records were considered good studies, with all quality criteria fulfilled.
Overall, the 48 selected studies were classified as good and acceptable quality and these were later used in subsequent data evaluations.
3.2. Dataset evaluation
Datasets were created through our database containing a summary of the selected in vitro and in vivo studies. The in vitro datasets comprised studies performed on germ cells, Leydig, Sertoli cells or human semen. In in vivo studies, 17 studies used the oral route of exposure, i.e. Ag NPs were administered via oral gavage or through food/water, 13 used intraperitoneal injection, 9 used intravenous injection, 1 used subdermal, 1 intratracheal, and 1 intratesticular administration routes, which were mostly conducted in rats (25 studies), mice (15 studies) and rabbits (2 studies).
The dose ranges used in these studies were quite wide, for example in the 17 oral administration studies, 0.015 to 500 mg/kg/bw doses were applied. The dose was generally linked with those selected in previous studies. Some of the studies pointed that the investigated doses were in the range between the lowest observable adverse effect level (LOAEL) and a no observable adverse effect level (NOAEL) (38,57), i.e., 125 mg/kg and 30 mg/kg, respectively, as suggested by a 90-day oral toxicity study, based on signs of liver toxicity (58).
In each study, the physicochemical properties of the used Ag NP was strictly reviewed, so that we could further identify the influence of these properties on the adverse outcome on the reproductive system. Ag NPs size range varied between 8.92 nm to 200 nm (Fig. 2).
Nine studies reported that the used Ag NPs were coated with polyvinyl pyrrolidone (PVP) or citrate. Moreover, some studies used Ag NPs produced via green method such as biogenic production in Bacillus funiculus, baker’s yeast (Saccharomyces cerevisiae) or V. opulus L. fruit extract. The studied Ag NPs were mostly spherically-shaped.
3.3. Potential molecular initiating events
Two mechanisms involved in the toxicity of Ag NPs have been proposed. First, silver ions are formed and released, mediated by the oxidation of the metallic silver core, inducing the formation of reactive oxygen species (ROS) (2). Silver ions interaction with enzymes and proteins containing thiol groups, such as metallothioneins or zinc-finger proteins, affecting cellular processes such as cellular respiration and antioxidant defense system, possibly resulting in cell death (12,13,22,59,60). Silver ions, indeed, are soft acid according to the HSAB principle, consequently they show high affinity towards soft bases and among them silver is highly affine for biomolecules containing thiols. However, it is important to go one-step back and explore the link between the oxidative stress and Ag NP dissolution to understand via which mechanism Ag NPs induce oxidative stress, in order to identify molecular initiating event(s).
Ag NP dissolution mechanisms are well described (9,61–63). The release of ions from Ag NPs has been shown to be an oxidation involving dissolved oxygen and protons. The reaction stoichiometry is as shown in the equation [1] (64),
$${2Ag}_{\left(s\right) }+ \frac{1}{2} {O}_{2\left(aq\right)}+2{H}_{\left(aq\right)}^{+}\leftrightarrow {2Ag}^{+}+ {H}_{2}O \left[1\right]$$
However, studies have shown that the oxidation of Ag NPs to Ag ions occurs through simple redox reactions that produce peroxide intermediates rather than a four-electron transfer process that directly reduces O2 to water (64,65). This is illustrated by the equation [2].
$${O}_{2}+{H}^{+} \underrightarrow{{Ag}^{0}} {Ag}^{+}+peroxide intermediates \underrightarrow{{Ag}^{0}} {Ag}^{+}+ {H}_{2}O \left[2\right]$$
Dissolution of Ag NPs to form Ag ions may accompany the decomposition of hydrogen peroxide under acidic conditions leading to the formation of hydroxyl radicals (63). The formation of hydroxyl radicals can go through a process similar to the Fenton reaction, in which Ag NPs act as a Fenton-like reagent (63). Under neutral and alkaline conditions, the reaction of Ag NPs with H2O2 generates oxygen instead of °OH radical (64). The pH of the environment plays a role in the Ag NPs dissolution, as illustrated by equation [3] (62).
$$Ag+ {H}_{2}{O}_{2}+ {H}^{+ }\to {Ag}^{+}+ \bullet OH+{H}_{2}O \left[3\right]$$
The ability of generating radicals has been reported across a variety of metal and metal oxide NPs, such as copper nanoparticles, zinc oxide nanoparticles, Ag NPs (66–68). Previous reports using electron spin resonance (ESR) coupled with spin trapping and spin labeling prove that free radicals are derived from the surface of Ag NPs (69,70).
In cellular system, Ag NPs are internalized within cells by endocytosis and are stored in endosomes, which then mature to lysosomes. Ag NPs undergo intracellular dissolution in the favor of low lysosomal pH, leading to Ag(I) species (11,16). This behavior is related to the so-called “Trojan horse” mechanism and leads to high Ag(I) concentrations in cellular compartments that Ag ions would otherwise not reach. Intracellular Ag(I) is a chemically reactive form of silver and shows remarkable affinity to zinc-finger domains of proteins, thiol-containing enzymes and molecules, mainly GSH and metallothionein. Binding to these ligands leads to the subsequent formation of intracellular Ag(I)-thiolate complexes (22,71,72). Such interaction affects the native domain structure of these proteins, which plays a role in maintaining the cellular homeostasis and antioxidant systems. As a consequence, it will influence their biological functions (13,60,71–73).
Due to the abovementioned mechanisms, since a MIE describes an initial point of interaction between stressors and the biomolecule, we propose that the impairment of intracellular SH-containing biomolecules can be defined as a MIE of the putative AOP described here. Indeed, the release of Ag(I) in solution, the consequent production of ROS together with the thiol-Ag ion complexation would activate such MIE and lead to a chain of intracellular consequences ultimately leading to reproductive toxicity.
In our MIE evaluations, we reviewed AOP207 (47), examining the reproductive toxicity study of Ag NPs in worms. This AOP focuses on identifying potential MIEs on Ag NPs induced reproductive toxicity in C. elegans (Table 2) and is still under development. The authors examine the question of how Ag NPs cause ROS production in C. elegans. They state that ROS can be formed on the surface of nanomaterials or that following the NP internalization endosomes are formed and ROS are produced by NADPH oxidase. They examine whether ROS arise directly from Ag NPs or indirectly through the action of NADPH oxidase. Finally, they identify NADPH oxidase as MIE, and reproduction failure as the outcome in C. elegans. However, any general correlation between the findings from C. elegans and in vitro and in vivo mammalian studies on the toxicity of Ag NPs is lacking (74).
3.4. Identification and selection of key events
AOP-Wiki was screened to identify already-existing KEs that could describe the biological events reported in the 48 selected studies. The result of this analysis and screening is reported in Additional file 1, an extract of which, reduced to the analysis of 4 articles, is presented in Table 1.
Disruption of SH-containing molecules (MIE), such as glutathione, can cause oxidative stress through disruption of the antioxidant system, as described in the section above. In our database, the most reported of the biological events is Ag NPs exposure triggering oxidative stress, which is described in 22 out of the 48 articles at both cellular and tissue level. It has been shown that accumulation of Ag NPs depleted cells from the molecular antioxidant GSH (39,75–77), and decreased super oxide dismutase (SOD) and catalase (CAT) activities (29,39,77–79), altered enzymatic oxidative defense system in male reproductive system (29,38,39,77,80,81) and lead to increased ROS levels in human sperm (82), in mouse Sertoli cells (15P-1) (83), in somatic Leydig (TM3) and Sertoli (TM4) cells (36,84) which eventually cause oxidative stress. In addition, mitochondrial damage due to the impairment of metallothionein (MIE) would result in oxidative stress by inhibition of electron transfer chain enzymes and perturbation of antioxidant system. Thus, it would increase mitochondrial ROS production, which may lead to mitochondrial damage including damage to respiratory chain and its membrane permeability. It has been shown that Ag NPs within the intracellular space has the potential to cause mitochondrial dysfunction by the depolarization of the mitochondrial membrane (84,85). Wang et al reported damaged mitochondria in the testis upon Ag NP exposure to Balb/c mice (86). These findings support the hypothesis in the biological plausibility perspective that Ag NPs interact with the thiol groups of the biomolecules, causing disruptions in the antioxidant system and thus triggering oxidative stress.
In addition to outlining the evidence supported by biological plausibility, there is also empirical evidence supporting this association in AOP-Wiki. AOP17 (87) has proposed a MIE similar to ours as the binding of electrophilic chemicals to SH(thiol)-group of proteins and /or to seleno-proteins, generating neurotoxicity. It is stated in AOP17 that soft metals like mercury binding to thiol/sulfhydryl/SH/SeH- groups results in structural modifications affecting the catalytic capacity of enzymes, and thereby reducing their capacity to neutralize ROS (88). The relationship of this MIE and oxidative stress is classified as moderate in quantitative manner. The same could occur with Ag ions. Therefore, it can be assumed that impairment of SH-containing molecules like glutathione and metallothionein can lead both mitochondrial damage (KE1a) and ROS production (KE1b), eventually result in oxidative stress (KE2) as shown in Fig. 3.
Table 1
Database summary of analysis 4 studies out of 48 selected studies
Ag NP
p-chem
|
Experimental model and mode of exposure
|
Exposure dose, sampling time
|
Biological/ toxicological
Endpoints
|
Actual measurements and methods
|
Result
|
Potential key event(s) (KE nr. from AOP-Wiki) a
|
Potential adverse outcomes (AOs) associated with the key event (AO nr. from AOP-Wikia
|
Ref
|
Size: 50 nm
Shape: Spherical
Hydrodynamic size :113.4 ± 12.1 nm
Zeta potential: −12.30 ± 0.4 mV
|
Adult male SD rats
Oral gavage
|
50 mg/kg bw, 3 months
|
Sperm Evaluation
|
Sperm Motility, Concentration, and Viability by eosin staining
|
Increased sperm morphological abnormalities
Decreased sperm concentrations motility
and viability
|
Decreased sperm quantity or quality in the adult, Decreased fertility (ID505, ID520)
|
Impaired fertility
(ID330 , ID406)
|
(38)
|
Oxidative Status
|
GSH level, CAT activity and lipid peroxidation MDA content in testicular tissue by commercial kits
|
Decreased CAT activity,
Increased MDA content,
Non-significant change in GSH level
|
Decreased protection against oxidative stress, Occurance oxidative stress (ID210, ID1112, ID1249, ID1538, ID1869)
Lipid Peroxidation (ID1445 or ID1511)
|
Oxidative Damage
(ID356)
|
Hormonal Assessment
|
Serum testosterone, LH and FSH by ELISA kit
|
Decreased testosterone, LH and FSH levels
|
Reduction, testosterone level (ID1613, ID1689, ID1612)
Reduced, Gonadotropins (ID1986)
|
Decreased sperm quantity or quality in the adult, Decreased fertility (ID505, ID520)
|
DNA damage
|
DNA strand breaks in testicular tissue by COMET assay
|
Increased DNA damage
|
Increased DNA damage (ID1194)
|
DNA Damage (ID1194)
|
Histopathological Examination
|
Testis and Seminiferous tubules by Hematoxylin and Eosin (HE) staining
|
Testis and Seminiferous tubules
Histological alterations
Necrotic spermatogonial cells
|
Testicular atrophy
(ID1506)
|
Male reproductive tract malformations (ID348)
Reduced, Reproductive Success (ID675)
|
Size:45 nm
PVP (< 1%)
Zeta potential:−20 mV
|
New Zealand White male rabbits
Intravenous
|
5 mM AgNP solution (0.6 mg/kg bw)
2 months (weekly)
|
Sperm Evaluation
|
Sperm Volume, Concentration, and Viability
|
Decreased sperm motility, concentration and volume
|
Decreased sperm quantity or quality in the adult, Decreased fertility (ID505, ID520)
|
Impaired fertility
(ID330 , ID406)
|
(75)
|
Oxidative Status
|
MDA content, NO concentration, CAT and GPX Activity in sperm and blood samples by commercial kits
|
Increased NO, and MDA content
Decreased CAT activity, and GSH level
|
Decreased protection against oxidative stress, Occurance oxidative stress (ID210, ID1112, ID1249, ID1538, ID1869)
Lipid Peroxidation (ID1445 or ID1511)
|
Oxidative Damage
(ID356)
|
Size:40 nm
Shape: Spherical
|
Adult male NMRI mice
Oral gavage
|
500 mg/kg bw with a time interval of 24 hr for 35 days
|
Oxidative Status
|
Total antioxidant capacity and Lipid peroxidation parameters
|
Decrease in the total antioxidant capacity, Increased MDA concent
|
Decreased protection against oxidative stress, Occurance oxidative stress (ID210, ID1112, ID1249, ID1538, ID1869)
Lipid Peroxidation (ID1445 or ID1511)
|
Oxidative Damage
(ID356)
|
(89)
|
Hormonal Assessment
|
Serum testosterone level by comercial kits
|
Decreased testosterone hormone
|
Reduction, testosterone level (ID1613, ID1689, ID1612)
|
Decreased sperm quantity or quality in the adult, Decreased fertility (ID505, ID520)
|
Histological parametres
|
Testis, the volume of interstitial tissue and seminiferous tubules
|
Decreased mean volume of testicular tissue and the volume of seminiferous tubules Decreased sperm density, mean number of spermatocytes, mean number of Sertoli cells
|
Testicular atrophy (ID1506)
|
Male reproductive tract malformations (ID348)
Reduced, Reproductive Success (ID675)
|
Size:100 nm
Shape: Spherical SSA : 7.5329m2/g
Zeta potential:− 18.9 mV
|
Male Rats
Sub dermal
|
10 and 50 mg/kg bw, 7 and 28 days
|
Sperm Evaluation
|
Sperm motility, velocity by HE staining
|
Decreased sperm motility and velocity
|
Decreased sperm quantity or quality in the adult, Decreased fertility (ID505, ID520)
|
Impaired fertility
|
(39)
|
Oxidative Status
|
MDA, GSH, CAT
|
Increased Lipid peroxidization Decreased SOD, CAT, GSH and total thiols
|
Decreased protection against oxidative stress, Occurance oxidative stress (ID210, ID1112, ID1249, ID1538, ID1869)
|
Oxidative Damage
(ID356)
|
Hormonal Assessment
|
Testosterone, LH and FSH
|
Decreased testosteron, LH and FSH levels (dose dependent
|
Reduction, testosterone level (ID1613, ID1689, ID1612)
Reduced, Gonadotropins (ID1986)
|
Decreased sperm quantity or quality in the adult, Decreased fertility (ID505, ID520)
|
Histological parametres
|
Cellular achitecture of testes and epididymis
|
Degenerative alterations in the cellular architecture of testes and epididymis
|
Testicular atrophy (ID1506)
|
Male reproductive tract malformations (ID348)
Reduced, Reproductive Success (ID675)
|
aAs several KEs and AOs are related to some of these cellular mechanisms, we indicate the title of only some of them, but the IDs of all of them are cited. |
Lipid peroxidation following Ag NP exposure has also been identified in our database. Lipid peroxidation byproducts including malondialdehyde (MDA), and Thiobarbituric acid reactive substances (TBARS) have been shown to be significantly increased in serum, testicular tissue or in reproductive cells exposed to Ag NPs (38,83,89–91). ROS-mediated lipid peroxidation is shown in at least two studies out of the 48 studies (29,75). Collodel et al (75) confirmed the correlation between excessive radical generation, lipid peroxidation, and damage to the sperm membrane. Evidence supporting the KER between oxidative stress and lipid peroxidation was also provided by AOP-Wiki and the relation (KER ID:1727) is classified as high by weight of evidence and quantitative understanding. Therefore, it can be postulated that oxidative stress leads to lipid peroxidation, as reported in Figure 3.
Ag NPs induced DNA damage is reported in some of the evaluated studies in in vitro germ cells, somatic cells (82,92) and in in vivo sperm samples and testicular tissues (38,76,79,80,93–96). According to AshaRani et al and Carlson et al, ROS formation/oxidative stress was suggested to be a key event in DNA damage induction (85,97).
Apoptosis is a widely observed response to Ag NP exposure, which is frequently reported through measurements of apoptosis-related proteins at the cellular level, or the tissue level with histopathological observations. At the cellular level, Ag NPs induce apoptosis in the mouse germ cell line C18-4 (98), in the mouse male-derived Sertoli cell line TM4 (84). It is also suggested that accumulated ROS lead to apoptosis as downstream event in somatic Leydig and Sertoli cells (36). Ag NPs induce expression of autophagy-related genes and activate signaling molecules involved in apoptosis (36). Ntera2 cells (NT2, human testicular embryonic carcinoma cell line) are affected by Ag NPs which cause DNA strand breaks, reduce the cell proliferation and trigger apoptosis and necrosis (94).
Moreover, several in vivo studies demonstrate alterations in apoptosis-related gene expressions, increased ratio of Bax/Bcl-2 expressions (77,80,86,99,100) and mitochondria-dependent intrinsic apoptotic pathway in testes (80,99). An extensive gene expression analysis conducted on 383 genes by microarray shows great changes in apoptosis-related genes and proteins (caspase3 and Myc). This analysis shows apoptosis-related changes of testis morphology and sperm production, with the evidence of apoptotic nuclei in spermatogonia and spermatocytes in the testis (80,86). The histopathology assessment of tubular cross-sections of seminiferous tubules provides evidence of increased number of apoptotic germ and somatic cells (30,101). Testicular sections in rats treated with Ag NPs show decrease and disturbance in the spermatogenic cells arrangements, atrophied seminiferous tubules with degenerative Sertoli cell, and depletion in Leydig cells (76). In addition, other studies using different target systems such as liver (23), colon (102), and endothelial cells (103) show that Ag NPs cause apoptosis in a p53-dependent process involving ROS and the c-Jun N-terminal kinase cascade, or via the IKK/NF-κB pathway. These results suggest the appropriateness of the KE ‘apoptosis’. Moreover, it is widely recognized that if cells fail to handle oxidative stress, then apoptosis will be triggered through downstream signaling pathways (104–107). Therefore, we chose to define apoptosis as a downstream event of mitochondrial damage, DNA damage and lipid peroxidation in the putative AOP that we propose (KE4).
In AOP-Wiki, apoptosis, DNA damage, and sperm count relation was evaluated in AOP 322 (108) (Fig. 4). In this network, DNA alkylation (MIE) cause subsequent key events as inadequate DNA repair, increased DNA strand breaks, increased apoptosis, and reduced sperm counts (Table 2). This AOP 322 is still under development, however, it provides a key sub-network that is possibly relevant to our proposed AOPs (Fig. 4). The AOP examining the reproductive toxicity study of Ag NPs in worms (47) classified PMK-1(P38 MAPK) activation, HIF-1 activation, mitochondrial damage, DNA damage, and apoptosis as key events (Fig. 4). They performed correlation analysis between each key events on their AOPs and it has been proved that there is significant positive correlations between the exposure concentration of Ag NPs, ROS formation, the expression of bli-3 (NADPH oxidase), and mitochondrial damage. The most significant negative correlations were observed between the concentration of Ag NPs, reproduction and DNA repair gene expressions in C. elegans (48). Our in vivo and in vitro data analysis is in good agreement with this AOP framework developed by Jeong et al which has similar key events as in our proposed putative AOP on mammalian models.
Significant alterations of serum and intratesticular testosterone levels was observed upon Ag NPs exposure, as reported in a number of studies from our database (38,39,78,79,100,109). According to Attia et al the significant decrease in the level of serum testosterone could be related with the adverse effects of Ag NPs in Leydig cells (78). Circulating testosterone levels depend on the steroidogenic capacity of individual Leydig cells and the total number of Leydig cells per testis (110). Leydig cell apoptosis causes the decrease in their number in the testis, which in turn affects testosterone level as shown in some studies (30,111,112) and further impact the spermatogenesis (113). Therefore, the relationship between apoptosis of Leydig cells and alterations of serum and intratesticular testosterone levels is consistent with established biological knowledge.
On the other hand, there are studies showing that low testosterone levels are associated with impaired cholesterol transport in damaged mitochondria of Leydig cells (114,115). Mitochondrial steroidogenic acute regulatory protein (StAR) or translocator protein (TSPO) are responsible for cholesterol transport from the outer to the inner mitochondrial membrane (116). In the inner membrane, cholesterol is converted into the Pregnenolone by CYP11a1 (116). Afterwards, 3β-Hydroxysteroid dehydorgenase (Hsd3b), 17β-Hydroxysteroid dehydorgenase (Hsd17b), CYP17A1 transform pregnenolone to testosterone (111,116). In our database, steroidogenesis perturbation by Ag NPs were shown in some reproductive toxicology studies. Garcia et al (117) reported no change in the expression level of StAR after Ag NP administration although they detected increased Cyp11a1 and Hsd3b1 expression levels, which were also verified through increased testosterone level in their study. However, Dziendzikowska et al (118) reported decreased expression level of Star, Cyp11a1, Hsd3b1 and Hsd17b3 in Wistar rats treated with Ag NPs. Zhang et al showed inhibited StAR, Hsd3b1, and Hsd17b transcription, which can negatively affect testosterone production in TM3 cells (36). As suggested by Dziendzikowska et al (119), impaired steroidogenesis is probably resulting from the interactions of Ag ions with the thiol groups present in the inner mitochondrial membrane (120,121). This concurs well with the MIE in the putative AOP that we propose.
In AOP-Wiki, impairment of steroidogenesis was investigating regarding its role in reproductive toxicity, and we identified AOP18 (122) that describes the AO impaired fertility following a MIE that is PPARα activation in Leydig cells (123). The development of this AOP relies on evidence collected from rodent models and incorporates human mechanistic and epidemiological data. The pathway comprise the activation of PPARα, followed by the disruption cholesterol transport in mitochondria, impairment of hormonal balance which leads to malformation of the reproductive tract in males. In their evidence assessment, the authors found a moderate relation between cholesterol transport in mitochondria and testosterone synthesis. It is stated that decreasing the amount of cholesterol inside the mitochondria (e. g by decreasing the expression of enzymes like StAR or TSOP) will result in a diminished amount of substrate for hormone (testosterone) synthesis. These results offer compelling evidence for the alteration of testosterone level in our pAOP. Therefore, we can assume that apoptosis in Leydig cells or its mitochondrial damage may lead to the endpoint of testosterone level reduction that we define as KE5 in our putative AOP.
In our database, after intravenous injection or oral route administration Ag NPs are shown to accumulate in the testes and are found in spermatids and ejaculated sperms, which suggest the likelihood that Ag NPs could pass through the blood-testis barrier (BTB) and eventually could impair the endocrine and reproductive functions (25,28,40,41,58,124–126). Arisha et al showed increased Ag NPs levels in testes, reduced expression of tight junction proteins (occludin, claudin-11, and tight junction protein 1) resulting in BTB permeability increase (99). They correlated these results with significantly reduced mRNA expression of hypothalamic GnRH1, testicular AR, and serum levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH) and testosterone concentrations. The authors conclude that decreased testosterone level, mainly due to unbalanced sexual hormones signaling and testicular damage, affects spermatogenesis. In a very recent study investigating steroidogenesis in rat hippocampus upon exposure to Ag NPs, the alterations in expression of Star, Hsd3b3, and Hsd17b1 genes, involved in steroid metabolism, is shown (119). Any interference with the normal functioning of the hypothalamic-pituitary-testicular axis can lead to reduced fertility, and if this interference persists, infertility could develop. Although these events do not occur in the reproductive organ itself but rather in the associated endocrine system, they contribute to Ag-NP reproductive toxicity and are complementary to all the other tested endpoints. Still, we chose not to include them in the current putative AO, which is focused exclusively on the male reproductive organ.
Sperm characteristics have a great importance in the prediction of fertility. Twenty-four out of the 48 identified studies analyzed sperm parameters after exposure to Ag-NPs, including sperm morphology, viability, motility, and DNA damage. Upon Ag NP exposure, abnormal sperm morphology, decreased sperm viability and motility, increased sperm DNA damage were reported in number of studies (38,75,77,79,86,90,109,127). The production of sperm requires a complex interaction between Sertoli, Leydig and germ cells. Any defect of these cells may prevent normal sperm production. Moreover, studies mostly emphasize that inadequate hormonal level may provoke low sperm quality and quantity (128). For example, Cavallin et al (95) showed that reduced sperm storage and reduced sperm transit time after Ag NPs exposure to rats may be due to increased testosterone concentration in the serum. In the AOP-Wiki database, the reduction of testosterone level is defined as an upstream event of decreased sperm quality. Therefore, we assigned decreased sperm quality and quantity (KE6) as a downstream event of reduction, testosterone level (KE5).
At the organ level, in vivo histopathological analysis of testis was conducted in 27 studies out of 48. Testis index and histological structure of testicular tissues, morphology, seminiferous tubule area, circumference, diameter and tubular degeneration/atrophy were the most studied parameters. A prominent atrophy of seminiferous tubules, thinning of the tubule wall, disorganization and vacuolization of germinal epithelium, and loss of spermatogenic cells in testis tissue of rats and mice exposed to Ag NPs are reported (39,77,80,84). In the interstitial tissue, Leydig cells are highly affected, presenting disrupted plasma membrane on extensive areas, with loss of cell organelles (80). Shehata et al showed significant reduction in the area, circumference and mean diameter of seminiferous tubules in Ag NP exposed rats (38). These results suggest that Ag NPs may affect the testicular structure and decrease reproductive success.
As a summary, impairment of intracellular SH-containing biomolecules may lead to mitochondrial damage and ROS accumulation (KE1) and lead to oxidative stress (KE2) which further provoke DNA damage and Lipid peroxidation (KE3). Intracellular perturbations may lead to apoptosis in Leydig cells. At the organ/organ system level, these perturbations result in altered testosterone level (KE5) and decreased sperm quality (KE6). All these biological events, which emerged from our literature analysis, lead to the putative AOP framework shown in Fig. 3.
3.5. Network of AOPs
Then, the pAOPs presented in Fig. 3 was used to tentatively build a network of AOPs for reproductive toxicity. MIEs and KEs involved in male reproduction impairment-related AOPs were extracted from AOP-Wiki and are listed in the Table 2.
Table 2
Male reproductive system AOPs on AOP-Wiki
AOP ID:
|
Title
|
KE
|
AO
|
Taxonomic Applicability
|
Ref
|
18
|
PPARα activation in utero leading to impaired fertility in males
|
MIE : Activation, PPARα
KE1 : Decrease, Steroidogenic acute regulatory protein (STAR)
KE2 : Reduction, Cholesterol transport in mitochondria
KE3 : Reduction, Testosterone synthesis in Leydig cells
KE4 : Reduction, Testosterone level
KE5 : Decrease, Translocator protein (TSPO)
|
Impaired, Fertility
Malformation, Male reproductive tract
|
Rattus norvegicus
Homo sapiens
Mus musculus
|
(122)
|
64
|
Glucocorticoid Receptor (GR) Mediated Adult Leydig Cell Dysfunction Leading to Decreased Male Fertility
|
MIE : Glucocorticoid Receptor Agonist, Activation
KE1 : Repressed expression of steroidogenic enzymes
KE2 : Increased apoptosis, decreased number of adult Leydig Cells
KE3 : Reduction, Testosterone synthesis in Leydig cells
KE4 : Reduction, testosterone level
KE5 : Decreased sperm quantity or quality in the adult, Decreased fertility
|
Impaired, Fertility
|
Rattus norvegicus
|
(129)
|
207
|
NADPH oxidase and P38 MAPK activation leading to reproductive failure in Caenorhabditis elegans
|
MIE : Activation, NADPH Oxidase
KE1 : ROS formation
KE2 : Increase, Oxidative Stress / Activation, PMK-1 P38 MAPK
KE3 : Activation, HIF-1
KE4 : Increased, DNA Damage-Repair
KE5 : Damaging, Mitochondria
KE6 : Apoptosis
|
Reproductive failure
|
Caenorhabditis elegans
|
(47)
|
208
|
Janus kinase (JAK)/Signal transducer and activator of transcription (STAT) and Transforming growth factor (TGF)-beta pathways activation leading to reproductive failure
|
KE1 : Activation, JAK/STAT pathway
KE2 : Activation, TGF-beta pathway
|
Reproductive failure
|
Caenorhabditis elegans
|
(51)
|
322
|
Alkylation of DNA leading to reduced sperm count
|
MIE : Alkylation, DNA
KE1 : Inadequate DNA repair
KE2 : Increase, DNA strand breaks
KE3 : Increase, Apoptosis
|
Reduce, Sperm count
|
No information
|
(108)
|
323
|
PPAR alpha Agonism Impairs Fish Reproduction
|
MIE : Activation, PPARα
KE1 : Decreased, cholesterol
KE2 : Decreased, 11KT
KE3 : Impaired, Spermatogenesis
KE4 : impaired, Fertility
|
No information
|
Teleost fish
|
(130)
|
444
|
Ionizing radiation leads to reduced reproduction in Eisenia fetida via reduced spermatogenesis and cocoon hatchability
|
MIE : Deposition of Energy
MIE : Increase in reactive oxygen and nitrogen species (RONS)
MIE : Increase, DNA damage
KE1 : Increased, Oxidative Stress
KE2 : Increase, Apoptosis
KE3 : Decreased spermatogenesis
KE4 : Decrease, Fecundity
KE5 : Decrease, Reproduction
|
Decrease, Population growth rate
|
No information
|
(131)
|
Some AOPs listed in this table are already discussed in the previous section. We observed that these AOPs share at least one common KE with our pAOPs. For example, oxidative stress, apoptosis, reduction of testosterone levels, decreased sperm quality events are shared by AOP 64, 207, 444. In Fig. 4, we interconnected these events with our pAOPs. While individual AOPs are likely to be activated by a limited number of reprotoxic compounds, interconnected AOPs that are linked by common key events of single AOPs are likely to represent more realistic descriptions of the complexity of disease pathophysiology (132).
The AOP network reported in Fig. 4 also shows the potential knowledge gaps in internal associations between KEs. Complete/pathway driven studies investigating the effects of impairment of SH-containing biomolecules and their role in male reproductive development are lacking. For establishing a solid quantitative linkage, mode of action framework analysis for reproductive toxicity is needed. This figure also could serve as a candidate list of MIE that could provide clues for experimental verifications for future studies.
3.6. Experimental methods for assessing the KEs
As suggested by Halappanavar et al (55), AOPs can be used as a tool in the design of testing strategies to support the safety assessment of nanomaterials. In this regard, our database identified the various in vitro endpoints, methods and assays used to measure the KEs in this pAOP (Table 3).
In studies examining oxidative stress, the intracellular level of ROS has often been evaluated using a fluorescent probe such as H2-DCF-DA (82,84,92). MDA content in ELISA methods can be used to detect lipid peroxidation, which is one of the main indicators of oxidative stress (77,81,89). With commercial kits, GSH levels and total antioxidant capacities can be measured, as well as analyzed at the level of antioxidant biomolecules and enzymes such as CAT, SOD (83,91,133). GSH and GSSG levels can also be determined biochemically in high performance liquid chromatography HPLC, capillary electrophoresis or microplates. Gene expressions of the antioxidant defense system can be measured by RT-PCR. Mitochondrial dysfunction can be measured by colorimetric assays such as 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) through assessment of mitochondrial membrane potential (MMP), mitochondrial ATP production, cytochrome c release, or mtDNA damage (84).
Although in the articles from our database DNA damage is evaluated using the comet assay, micronucleus assay and DNA fragmentation, a roadmap for testing DNA damage caused by nanomaterials was recently proposed by Elespuru et al (134). It includes the use of an in vitro gene mutation assay (OECD TG476 (135), HPRT or TG490 (136), mouse lymphoma TK ± assay) and a chromosomal damage assay (OECD TG487 (137) in vitro micronucleus assay or TG473 (138) chromosomal aberration assay). Eventually, optional assays are proposed, both in vitro (comet assay) and in vivo (comet assay, OECD TG489 (139); transgenic rodent gene mutation assay TG488 (140); erythrocyte micronucleus assay, TG474 (141); bone marrow chromosomal aberration test, TG 475 (142)). Other non-guideline test methods to measure the DNA damage also exist although they are not discussed in the roadmap by Elespuru et al (134). For instance the detection of DNA repair proteins such as H2AX, 53BP1 or XRCC2 can be used, or high performance liquid chromatography (HPLC) coupled to tandem mass spectrometry (MS/MS) that quantifies very low levels of oxidative lesions to DNA (143).
In articles in our database, apoptosis was evaluated by western blot, qPCR, TUNEL Assay or Flow Cytometry (80,84,98,144). Other methods can also be used, including annexin V-FITC probes, with analysis of the relative percentage of Annexin V-FITC-positive/PI-negative cells by flow cytometry. The alteration of procaspases 7 and 3, Caspase-3 and caspase-9 activity, as well as the cleavage of PARP can be determined by western blotting or RT-PCR.
The OECD TG 456 (145) is a validated test guideline for in vitro screening of the effect of chemicals on steroidogenesis, specifically the production of 17ß-estradiol (E2) and testosterone. In vitro testosterone synthesis in Leydig cells can be measured by P450scc, StAR, Hsd3b, and Cyp17a1 gene expression or indirectly by testosterone radioimmunoassay or analytical methods such as LC-MS or by isotope-dilution gas chromatography-mass spectrometry in serum (146,147). Sperm assessment includes the evaluation of sperm count and concentration (haemocytometer, automated image-based system), morphology and motility (microscope, automated image-based system) and viability (for example propidium iodide staining of necrotic cells, TUNEL assay staining apoptotic cells).
Models that can be used as alternatives to animal experimentation for assessing this putative AOP on male reproductive function, i.e., in vitro models of Leydig cells, Sertoli cells, Sertoli-germ cell cocultures, as well as methods to prepare testicular organ and tissue culture systems can be found in the Database Service on Alternative Methods to animal experimentation (DB-ALM) (148). Data generated by alternative methods or in vivo testing can be integrated in quantitative AOPs and can be validated for future studies.
Another purpose of our research in developing this pAOP was actually to observe how physicochemical properties of NPs influence the key events and their relationships. From the selected articles that were analysed, we identified Ag NP size, agglomeration state, surface coating and tendency to dissolve as key physico-chemical parameters that could influence their toxicity (Additional file 1). However, it is difficult to reach a definite conclusion, as the physicochemical characterization of Ag NPs in the considered articles is too diverse and sometimes lacks precision. It seems clear from these studies that controlling ion release could diminish the hazard potential of Ag NPs with respect to SbD approaches (16). Therefore, it is highly recommended to analyze Ag ion release systematically in the published articles. Since ion release is higher when the nanoparticle is smaller and when the nanoparticle surface is uncoated or coated with a ligand that tends to desorb, we consider that both the size and surface coating are important parameters that influence Ag NP hazard potential, as previously suggested in studies related to other organs (13,149,150).
Table 3
Summary information of the biological endpoints measured in the reproductive system toxicology
KE
|
Biological events/measurement
|
Methods
|
Ref
|
Cell Viability
|
|
MTT, MTS, LDH, CCK-8 assay
|
(83,84,92,94,98)
|
Oxidative stress
|
ROS Production
|
H2DCFDA
|
(82,84,92)
|
Lipid peroxidation,
MDA content
In testicular tissue and cells
|
ELISA, Western Blot, qRT-PCR
|
(76,77,81,83,89,96)
|
Enzymatic/non-enzymatic antioxidants
GPX, CAT, SOD, TBARS, TAOC, GSH
in Sperm, Seminal Plasma, and Blood, serum, testicular tissue homogenate and in vitro cells
|
ELISA, Western Blot, qRT-PCR
|
(38,75,76,83,91,133)
|
DNA damage
|
DNA Strand breaks,
In cells and the testis tissue
|
Comet Assay, DNA microarray analysis,
DNA fragmentation
|
(38,75,76,83,94,96)
|
Mitochondrial damage
|
Cell metabolic activity,
|
MTT assay,
|
(98)
|
Mitochondrial membrane potential
|
Quantitative method uses a tetraphenylphosphonium ion (TPP+)-sensitive electrode, or by fluorimetric methods
|
(84)
|
Apoptosis, in germ Sertoli and Leydig cell
|
P53, BAX, bcl-2 gene expressions or cell apoptosis
|
Western Blot, qPCR, TUNEL Assay Flow Cytometry
|
(77,80,84,98,99,144)
|
Reproductive hormone levels
|
Hormonal Assessment
Serum Testosterone, LH and FSH,
P450scc, StAR, Hsd3b, and Cyp17a1 gene expression
|
Radioimmunoassay (145), ELISA, qRT-PCR, chemiluminescent protein immunoassay
|
(38,76,78,86,89,118,151)
|
Sperm evaluations
|
Morphology, Motility, Concentration, Count, Viability, Plasma membrane intergrity
|
Eosin/nigrosine staining, hemocytometer, Microscope
|
(38,79,82,90,126,144,152)
|
Acrosome status
|
Fluorescence assessment (eg.chlortetracycline fluorescence assay)
|
(57,126,152)
|
Sperm DNA integrity
|
Toluidine blue staining, Aniline blue staining, Acridine orange staining Eosin–nigrosine-staining
|
(29,77,90,109,153)
|
Sperm DNA damage
|
Comet assay
|
(96,153)
|
The mitochondrial activity
|
Activity of cytochrome c oxidase
|
(57,95)
|
Spermatogenesis
|
The transcript expression of Gnrh1, Ar, Cyp11a1, Hsd3b1, Hsd17b3, Srd5a1, Cyp19a1, Star by qPCR
Intratesticular steroid metabolism enzyme protein level aromatase (Aro) and 5α-reductase type 1 (Srd5a1) measurements.
|
(95,118)
|
Determination of silver concentration in organs
|
|
ICPMS, UV/vis proton spectrophotometery
|
(29,100,154)
|
Histopathological Examination
|
The seminiferous tubules area, circumference, and diameter,
Testis index and histological structure of testicular tissues
|
Hematoxylin and Eosin light microscope Transmission electron microscopy (TEM)
|
(38,86,90,109)
|