Knockout of ho-1 protects the striatum from ferrous iron-induced injury in a male-specific manner in mice

Men have worse survival than premenopausal women after intracerebral hemorrhage (ICH). After ICH, overproduction of iron associated with induction of heme oxygenase-1 (HO-1) in brain was observed. Rodent ICH model using ferrous citrate (FC)-infusion into the striatum to simulate iron overload, showed a higher degree of injury severity in males than in females. However, the participation of HO-1 in sex-differences of iron-induced brain injury remains unknown. The present results showed a higher level of HO-1 expression associated with more severe injury in males compared with females after FC-infusion. Estradiol (E2) contributed to lower levels of FC-induced HO-1 expression in females compared with males. Heterozygote ho-1 KO decreased the levels of FC-induced injury severity, histological lesions, behavioral deficits, autophagy and autophagic cell death in the striatum of males but not in females. Moreover, ho-1 deficiency enhanced the neuroprotection by E2 only in males. These results suggested that over induction of HO-1 plays a harmful role in FC-induced brain injury in a male-specific manner. Suppression of HO-1 combined with E2 exhibits a synergistic effect on neuroprotection against FC-induced striatal injury in males. These findings open up the prospect for male-specific neuroprotection targeting HO-1 suppression for patients suffering from striatal iron overload.

Scientific RepoRts | 6:26358 | DOI: 10.1038/srep26358 in a sex-dimorphic manner. The present study was designed to test the hypothesis that ferrous iron causes HO-1 over induction and suppression of HO-1 protects the striatum from iron overload in a male-specific manner.

Sex differences in the FC-induced HO-1 expression and striatal injury.
To know whether there is a correlation between the level of HO-1 induction and the sex-differences of striatal injury caused by iron overload, the levels of HO-1, injury severity, histological lesions and behavioral deficits were evaluated in both male and female mice after FC-infusion. The results showed that FC infusion increased the HO-1 protein levels in the striata of male and female mice by 4.9-fold and 3. 5-fold, respectively (Fig. 1a). With a similar trend, FC infusion increased the levels of SBDP 145/150, which depicts injury severity, by 8.8-fold vs 4.0-fold (Fig. 1b), the lesion ratio by 16.3-fold vs 7.9-fold ( Fig. 1c) and the forelimb use asymmetry score by 9.6-fold vs 7.4-fold ( Fig. 1d) in male vs female mice, respectively. The iron deposition, which is confirmed by Prussian blue assay, after FC infusion in the striatum of male mouse was more than that in female as shown in the supplementary information (Fig. S1). The images by HE stain showed that the histological lesion in the striatum of male was also severer than that of female after FC infusion as shown in the supplementary information (Fig. S2).

Role of E 2 in sex differences of HO-1 induction and spectrin cleavage after FC infusion.
To examine whether endogenous E 2 contributes to the lower levels of FC-induced HO-1 expression and injury severity in females, the levels of HO-1 and cleavage of α -II spectrin were compared between FC-infused castrated mice with and without E 2 implantation. The results showed that orchiectomy did not change the level of HO-1 induction by FC in males, but ovariectomy increased the level of HO-1 induction by FC infusion in females. Orchiectomy diminished the sex-differences in FC-induced HO-1 expression and injury severity. E 2 implantation decreased the protein levels of HO-1 in castrated males and ovariectomized females after FC-infusion (Fig. 2a). Similarly, orchiectomy did not change the level of FC-induced cleavage of spectrin in males, but increased the level of FC-induced cleavage of spectrin in female mice. E 2 implantation decreased the levels of FC-induced cleavage of spectrin to SBDP 145/150 in both castrated males and females (Fig. 2b).
To address whether E 2 inhibits the FC-induced HO-1 expression at a translational level or transcriptional level, the levels of mRNA and protein were compared between castrated mice with and without E 2 implantation. The results showed that, after orchiectomy/ovariectomy, no sex difference in protein or mRNA levels of HO-1 induction by FC-infusion was observed. In castrated males, exogenous E 2 decreased the levels of HO-1 protein and mRNA by 67.7% and 69.9%, respectively. In ovariectomized females, E 2 decreased the HO-1 protein level by 35.0% and mRNA by 7.7% (Fig. 3a,b).
Effect of ho-1 deficiency on FC-induced striatal injury and autophagy. As shown in Fig. 4a, the levels of HO-1 in both male and female ho-1 +/− mice were significantly lower than those in ho-1 +/+ mice, respectively. Similar to C57BL/6 mice, FC-induced a higher level of HO-1 in male than in female ho-1 +/+ mice. While no sex-difference in FC-induced HO-1 expression in ho-1 +/− mice was observed (Fig. 4a).   To elucidate the effect of heterozygote ho-1 KO on FC-induced striatal injury, cleavage of α -II spectrin, histological lesions, and behavioral deficits were compared between ho-1 +/− and ho-1 +/+ mice after FC infusion. Similar to the result from C57B/6 mice, FC infusion caused higher levels of SBDP 145/150 (Fig. 4b), lesion ratios (Fig. 4c), and forelimb use asymmetry scores (Fig. 4d) in ho-1 +/+ males than those in ho-1 +/+ females. Heterozygous KO of the ho-1 gene decreased the levels of FC-induced spectrin cleavage, histological lesions, and behavioral deficits in male but not in female mice, and diminished the sex differences in FC-induced striatal injury ( Fig. 4b-d).
We next investigated whether heterozygote ho-1 KO affected the FC-induced autophagy and autophagic cell death, the level of LC3-II and the number of TUNEL(+ ) BECN1 immunoreactive cells were compared between striata of ho-1 +/− and ho-1 +/+ mice after FC-infusion, respectively. The results showed that heterozygous KO of the ho-1 gene decreased both BECN1 immunoreactivity and the number of TUNEL(+ ) cells in the striatum of male mice with FC infusion (Fig. 5a). The representative images were taken in the enclosed areas (striata) as indicated in the supplementary information (Fig. S2). The quantitative result showed that, without FC-infusion, heterozygous KO of the ho-1 gene did not change the number of nuclei in both male and female mice. FC-infusion decreased the number of nuclei in males and heterozygous KO of ho-1 gene diminished the FC-induced nucleus loss in males but not in females (Fig. 5b). FC-infusion induced higher levels of DNA fragmentation (TUNEL positive), BECN1 immunoreactivity, and TUNEL(+ ) BECN1 immunoreactivity in males than in females (Fig. 5c-e). Heterozygous KO of the ho-1 gene decreased the number of TUNEL(+ ) cells, BECN1 immunoreactive cells and TUNEL(+ ) BECN1 immunoreactive cells after FC-infusion in males, but not in females. Similarly, FC-infusion caused a higher ratio of LC3-II to LC3-I, which depicts level of autophagy in males compared to females, whereas heterozygote KO of ho-1 gene decreased the level of FC-induced autophagy in males but not in females. Interestingly, without FC-infusion, the basal level of LC3-II was higher in males than in females and heterozygote ho-1 KO diminished the sex-differences in the level of constitutive autophagy in intact mice (Fig. 5f).

Discussion
The present study demonstrated that a higher level of HO-1 associated with higher levels of injury severity, histological lesions and behavioral deficits in males than in females was induced by striatal FC-infusion, which simulates iron overload in the striatum after ICH. E 2 contributes to the lower levels of FC-induced HO-1 expression and injury severity in females than in males. Heterozygote KO ho-1 gene diminished the FC-induced striatal injury in males but not in females. In addition, heterozygote KO ho-1 gene simulates and enhances the neuroprotective effect of E 2 on FC-induced striatal injury. These results suggested that the high level of FC-induced HO-1 in males may exaggerate the FC-induced striatal injury, therefore HO-1 suppression diminished the FC-induced striatal injury and favored the outcome of males after iron overload. Our study, for the first time, explored the sex dimorphic effect of HO-1 suppression on the FC-induced striatal injury. These findings open up the prospect for a male-specific neuroprotection targeting HO-1 suppression for patients suffering from striatal iron overload.
Understanding of the sex difference in neurological recovery and responses to therapeutics may help the successive preclinical translation. There is growing literature examining sex differences in stroke epidemiology, however conflicting evidences of sex differences in outcome exist after stroke. Previous study reported that, without examining the effect of sex by stroke subtype, women aged 55 to 84 years had lower mortality than men, while women aged more than 85 years had 15% higher stroke mortality than men 2 . Another report indicated that, for younger patients after ICH, female sex was protective; however, at ages greater than 60 years, female sex was a risk factor for discharge to hospice or death. Their statistical analysis about the interaction between gender and age demonstrated significantly strong correlation with early outcome after ICH 17 . On the other hand, Dehlendorff et al. reported that, after the age of 60 years, elderly women were affected more severely than men with ischemic stroke but not with hemorrhagic stroke 18 . Recently, the report from heart disease and stroke statistics-2016 updates indicated that women have greater levels of disability than men after stroke 19 . Usually, women are older than men when they have a stroke, and severity increases with age. Moreover, the numbers of older women were higher than men in the population, the clinical trial enrollment with analyzing stroke types or age factor separately may decrease the inconsistency in the sex-difference in the severity after stroke. In the present study, we used FC-infusion mouse model to simulate accumulation of ferrous iron after hemorrhage. The result showed that FC-infusion induced a higher degree of striatal injury in male mice than that in age-matched females indicating FC-infusion model simplifies the detrimental factor and excludes the effect of age factor after hemorrhagic stroke. This model is proper for studying sex-dimorphic striatal injury due to iron overload that simulates hemorrhagic stroke.
The ho-1 gene is exquisitely sensitive to induction by a wide range of pro-oxidant and other stressors including ICH 20 . After hemorrhage, clot lysis and iron resulting from hemoglobin degradation play important roles in ICH-induced brain injury 21 . Brain iron accumulation may induce neuronal damage even after it has become bound to ferritin because iron can be released in its ferrous form under the acidic conditions 22 . Iron toxicity is largely based on Fenton chemistry where iron reacts with reactive oxygen intermediates to produce highly reactive free radical species such as the hydroxyl radical (OH − ) 23 . A previous report indicated that deferoxamine (an iron chelator) reduced the number of HO-1 (p < 0.01) positive cells in the ipsilateral basal ganglia in a rat model of ICH by intracaudate injection of autologous blood 24 , suggesting iron overload contributes to the HO-1 induction after ICH. In the present study, castration diminished the sex-difference in FC-induced HO-1 expression. Moreover, castration increased, while E 2 implantation decreased, the levels of HO-1 induction (Fig. 2a). This suggested that the suppressive effect of E 2 on HO-1 induction contributes to the lower level of HO-1 induction associated with injury severity caused by FC-infusion in females compared with males.
The effect and mechanism underlying the effect of E 2 on HO-1 induction remains controversial. A previous report indicated that E 2 treatment led to Nrf2 dissociation from Keap1 in the cytoplasm, and then translocated into nucleus, with a significant increase in HO-1 expression in homocysteine-treated cells 25 . Another study demonstrated that the protective effects of E 2 in male rats were ERα -independent and might be associated with HO-1 inhibition 26 . In the present results, E 2 significantly inhibited HO-1 induction by FC infusion in the striatum of male mice at both levels of mRNA and protein, while E 2 inhibited FC-induced HO-1 expression at the level of protein but not at an mRNA level in females (Fig. 3). This result suggested an ERα -independent mechanism underlying the suppression of HO-1 expression by E 2 in the striatum of male mice, because the expression level of ERα mRNA and protein and the number of ERα immunoreactive cells in the striatum were higher in female than in male brains 27 . However, further investigation is needed to understand how E 2 decreases FC-induced HO-1 expression. HO-1 is an enzyme exhibiting both beneficial and harmful effects, depending on the different experimental models 28 . Its beneficial effects have been related to the catabolism of the prooxidant heme to biliverdin and bilirubin, which are potent antioxidants 29,30 , if produced in excess, unconjugated bilirubin becomes neurotoxic through multiple mechanisms involving the disruption of cell membrane structure, the reduction of mitochondrial transmembrane potential and the activation of the apoptotic cascade 31,32 . While a massive HO-1 activation may metabolize heme into free iron protoporphyrin, it then releases equimolar amounts of ferrous iron 33,34 , which catalyzes reactions that generate ROS. Iron chelators decreased microglia activation in the ipsilateral hemisphere after hemin and quinolinic acid injection, strongly suggesting that Fe 2+ may be implicated in deleterious HO-1 effects 35 . Previous reports indicated that a HO-1 inhibitor, SnPP, attenuated edema in a hemoglobin-induced model of ICH 36 and HO-1 inhibition by tin-mesoporphyrin IX (SnMP) protection against neuronal loss in a rabbit model of autologous blood injection 37 . In addition, SnMP-treatment also reduced edema development following experimental ICH in pigs 38 . These results suggest a harmful role of HO-1 in hemorrhagic stroke. Moreover, deletion of the ho-1 gene results in diminution of brain tissue injury in a male mouse model of ICH by collagenase injection in the caudate putamen 39 . In the hemorrhagic brain of mice, HO-1 expression is mainly observed in microglia/macrophages around the hematoma region. Notably, ho-1 deletion is also associated with a decreased number of activated microglia/macrophages in thrombin-induced brain injury, suggesting that HO-1 supports activated microglia survival, which may exacerbate striatal injury 40 . In the present results, a significantly higher level of FC-induced HO-1 expression in the striatum was associated with a higher severity of striatal injury of intact males than that in intact females after FC-infusion (Fig. 1a). Furthermore, heterozygote KO ho-1 gene diminished the FC-induced striatal injury as well as the FC-induced autophagy and autophagic cell death in males but not in females, implicating a harmful role for massive HO-1 induction in males. While the moderate level of HO-1 induction after FC-infusion in females may help with hormesis under a redox microenvironment.
Basal HO-1 expression is maintained at a low level in the normal brain and is restricted to small groups of scattered neurons and neuroglia 12 . Induction of glial ho-1 may lead to the accumulation of iron 41 , which may exacerbate intracellular oxidative stress and subsequent injury by promoting free radical generation within 42 . The activity of heme oxygenase has also been shown to enhance vulnerability in various experimental models of neural injury and disease 43,39,44 . A previous report indicated that over-expression of the ho-1 gene promotes mitochondrial membrane damage, unregulated iron deposition, and macroautophagy in cultured astroglia 45 . Our recent report showed that iron-induced striatal injury is alleviated by inhibition of autophagy in a male-specific manner 46 . Whether the deleterious HO-1 activation effect in males on iron toxicity is linked to autophagy remains unknown. In the present study, we examined whether autophagy inhibition mediates the protective effect of ho-1 deficiency on FC-induced striatal injury. Interestingly, we found that, without FC-infusion, the basal level of constitutive autophagy was higher in males than in females and ho-1deficiency diminished the sex-difference in the level of constitutive autophagy (Fig. 5f). This result suggested that the suppressive effect of E 2 on the basal level of autophagy is HO-1 dependent. On the other hand, after FC-infusion, both the amount of autophagic cell death and the level of LC3-II in ho-1 +/+ males were higher than those in ho-1 +/− males, respectively. Although the interaction between HO-1 and autophagy remained obscure, these results implied that HO-1 may enhance vulnerability in FC-induced autophagy and autophagic cell death. Therefore, ho-1 deficiency conferred protection, at least partially, by diminishing the FC-induced autophagy and autophagic cell death.
Estrogen is both a natural neuroprotectant and a potential therapeutic agent for cerebrovascular disease. Pretreatment with E 2 attenuates brain edema after ICH in male mice 47 . In addition, estrogen reduces iron-induced brain edema in vivo and neuronal death in vitro 48 . Several mechanisms may contribute to the beneficial effect of estrogen on brain injury. A previous report showed that administration of E 2 to male (but not female) rats significantly reduced brain edema, neurological deficits, and the level of HO-1 induced by ICH 26,49 . This effect of E 2 in males might be ER-α -independent and might be associated with HO-1 inhibition 26 . Another report indicated that E 2 protects against light-induced retinal damage via its antioxidative effect 50 . In the present results, no significant effect of ho-1 deficiency on E 2 neuroprotection was observed in the female group (Fig. 6). While, in male mice, ho-1 deficiency per se slightly simulated the protective effect of E 2 on FC-induced striatal injury, suggesting HO-1 suppression partially mediates the neuroprotective effect of E 2 . Moreover, ho-1 deficiency further enhanced the protective effect of E 2 against FC-induced striatal injury in males but not in females, implicating the synergistic protection of HO-1 suppression with E 2 may be ERα -independent because the level of ERα in the striatum was higher in females than in males.
In conclusion, HO-1 plays a sex dimorphic role in FC-induced brain injury and HO-1 silencing has a sex-specific benefit for iron-induced striatal injury only in males because they lack the endogenous protection conferred by estradiol. These findings open up the prospect for a male-specific injury prevention targeting HO-1 inhibition for patients suffering from acute iron overload caused by ICH.

Materials and Methods
Animals. A total of 120 age-matched (12-week-old) male (♂ ) and female (♀ ) C57BL/6 mice [purchased from the National Laboratory Animal Center, Taipei, Taiwan] were used for studying the sex-differences in FC-induced HO-1 expression and injury severity. To study the contribution of E 2 to the sex difference in FC-induced HO-1 expression in the striatum, C57BL/6 mice implanted with estradiol (E 2 ) after castration were used. For female mice, ovariectomy involves a small incision in both flanks after which the ovaries are removed. In male mice, the skin and body wall caudal to the penis were incised with a scissor, and the testicles were retracted through the incision. The exposed testicular artery and vein were double-ligated and cut 51 . A silastic tube (2 mm inner diameter, 20 mm in length) containing 0.38 mM E 2 (Sigma-Aldrich, E8515) was implanted subcutaneously 24 h before FC infusion. Three μ l of FC (1 nmol/μ l) (Sigma-Aldrich, St. Louis, USA Sigma) were infused into the right striatum (coordinates: 0.2 mm anterior, 2.5 mm lateral, and 3.5 mm ventral to the bregma) using a microinfusion pump [CMA Microdialysis] at a rate of 1 μ l/min. Sham-operated mice were used as the control group. Two days Scientific RepoRts | 6:26358 | DOI: 10.1038/srep26358 after FC infusion, the forelimb use asymmetry test was performed to evaluate the FC-induced behavioral deficits. Then, the brain tissue containing the striatum was sampled and frozen for western blot analysis or was fixed and sectioned in 10 μ m thicknesses for the evaluation of histological lesions or immunohistochemical detection. In addition, a total of 144 age-matched (12-week-old) male (♂ ) and female (♀ ) heterozygous ho-1 KO (ho-1 +/− ) mice [breeding pairs of ho-1 heterozygous mice was derived from Dr. Shaw-Fang Yet, Harvard University via Brigham and Women's Hospital, Boston, MA, USA] were used for investigating the role of HO-1 in sex differences of FC-induced striatal injury. Heterozygous ho-1 KO (ho-1 +/− ) mice were used because ho-1 −/− mice were neonatally lethal. No animals were excluded. All operations were performed under anesthesia with Zoletil 50 (1 ml/kg B.W. intraperitoneally) and were approved by the Kaohsiung Medical University Committee for the Use of Experimental Animals. All experiments were performed in accordance with the approved guidelines.
Quantification of histological lesions. The paraffin-embedded tissue was serially sectioned into sections of 10 μ m in thickness. After hematoxylin and eosin (H&E) staining, the level of histological lesions in every fifteenth section containing the striatum was analyzed using Image-pro Plus software (Universal Imaging Corp.) according to the staining intensity of the striatal area. The histological lesion ratio was estimated by dividing the hemispheric volume of the striatum on the ipsilateral side by that on the contralateral side. The iron accumulation was confirmed by Prussian blue assay on brain sections from mice infused with FC. Real-time polymerase chain reaction. Total RNA from brain tissue containing the striatum was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA). Complementary DNA was transcribed at 42 °C for 60 min using SuperScript II RNase H-Reverse Transcriptase with oligo and random primers (Promega Corporation, Madison, Wis) in a reaction mixture containing 50 mM Tris-HCl, 50 mM KCl, 10 mM MgCl2, 10 mM dithiothreitol, 0.5 mM spermidine, and 1 mM deoxynucleotide triphosphates. Primers for HO-1 and 18S ribosomal RNA were designed as the following sequences, respectively: 5′ -AGGTGTCCAGGGAAGGCTTTA-3′ (forward) and 5′ -TAATGCCTTCCCTGGACACCT-3′ (reverse) and 5′ -CGCAGCTAGGAATAATGGAATAGG-3′ (forward) and 5′ -CATGGCCTCAGTTCCGAAA-3′ (reverse). Real-time PCR was performed in a 25 μ L reaction mixture containing 12.5 μ L of 2x SYBR Green PCR Master Mix (Applied Biosystems), 50 ng complementary DNA, and either 300 nM primers for HO-1 or 100 nM primers for 18S rRNA. Samples were analyzed in triplicate using the following thermocycling conditions: 50 °C for 2 min and 95 °C for 10 min, denaturation at 95 °C for 15 s, annealing at 60 °C for 1 min for 40 amplification cycles in total.
Immunostaining and TUNEL staining. The tissue sections were deparaffinized and rehydrated by immersing in xylene followed by a series of immersions in graded ethanol and double distilled water. The sections were then washed in 0.1 M PBS for 10 min and incubated for 30 min at 37 °C with proteinase K solution, nuclease free treatment, followed by quenching endogenous peroxidase with 3% H 2 O 2 in methanol for 20 min. For antigen retrieval, the sections were placed in 0.1 M citrate acid (pH 6.0) and irradiated with 350 W microwaves for 5 mins. Then sections were subsequently incubated with blocking solution containing 2% Bovine Serum Albumin (BSA, Sigma chemical CO.), 10% Normal Goat Serum (NGS, Jackson ImmunoResearch Laboratories, Inc., West Grove, USA) and 1% Triton X-100 for 30 mins at room temperature. Mouse anti-BECN1 antibodies Automated characterization and quantification of TUNEL(+) BECN1 immunoreactive cells. Images were captured with a fluorescent scanning microscope, followed by an in situ quantification using microscopy-based multicolor tissue cytometry (MMTC). In brief, tissue sections were scanned using a TissueFAXS ® cytometer (TissueGnostics, Vienna, Austria). Images of the entire tissue sections were captured at 20× amplification. Using the system's built-in autofocus, the field of striatum was figured-out to estimate the area of the striatum. Analysis of immunofluorescent tissue sections was acquired by the TissueQuest ® analysis software (TissueGnostics) that identified TUNEL(+ ) BECN1 immunoreactive cells using a proprietary search algorithm that first identified DAPI-labelled nuclei and then measured the mean relative fluorescence intensity around each nucleus. The mean relative fluorescent intensity turned out to be the best parameter to visualize BECN1 immunoreactive populations and was used throughout the whole study to distinguish positive and negative cell populations. The fluorescence reactivity determined by TissueQuest ® for the negative control samples defined the level of unspecific staining reactivity and allowed determination of the individual cut-off value for each sample. Only objects exceeding the fluorescence intensity in the respective channel were accepted as specifically stained and therefore as marker-positive cells. The amount of DAPI(+ ) nuclei and the number of TUNEL(+ ) or BECN1 immunoreactive cells for each sample were analyzed simultaneously 55 .
Statistical analysis. FC-induced injury in the striata from males and females was compared using two-way ANOVA followed by a Scheffé post hoc test. Data on the effects of ho-1 knockout on FC-induced injury severity, histological lesions, behavioral deficits, autophagy and autophagic cell death were analyzed using a multi-way ANOVA to determine the effect of each factor and the interaction between two factors. Significance was accepted at p < 0.05.