Genetic analysis of a novel antioxidant multi-target iron chelator, M30 protecting against chemotherapy-induced alopecia in mice

CTX and the antioxidant, M30 were purchased as injectable commercial products (Sigma-Aldrich, USA). CTX and M30 were freshly dissolved in phosphate buffer saline (PBS; Gibco, USA) before injection.

Six-weeks-old female C57BL/6 mice were obtained from Orient Bio Inc. (Korea). The mice were maintained in temperature-controlled clean racks with a 12-h light/dark cycle. The mice were allowed to acclimatize for 1 week before the start of the experiment. All experiments were performed in accordance with the institutional guidelines of the Korea Institute of Radiological & Medical Sciences (KIRAMS).

The mice were depilated by shaving, followed by waxing using wax strips (Veet, USA); the skin was washed with PBS before experiments. At 9 days after depilation, the alopecia models were generated by a single intraperitoneal injection of CTX (120 mg/kg body weight) that was freshly prepared in PBS as previously described [21]. Before intraperitoneal injection of CTX, one group of mice (n = 5) was subcutaneously implanted with Alzet mini-osmotic pump (model 2004, DURECT, Cupertino, CA) containing M30 in PBS with a delivery rate of 1 mg/d/kg. The other normal (n = 5) and CTX treatment groups (n = 5) underwent sham operations without M30 supplement.

For the euthanization of mice, CO₂ was injected at a rate of 10–30% per minute, that gradually fills the euthanasia chamber. Euthanization was performed by trained member. Then full-skin thickness samples of skin tissue were excised using scissors and immediately fixed using 4% paraformaldehyde in PBS for overnight at 4 °C. After fixation, the samples were incubated with 30% sucrose in PBS for 24 h at 4 °C. Then, the skin wasr embedded in O.C.T compound (Fisher Scientific, Pittsburg, USA) and frozen in liquid nitrogen. After freezing, the samples were stored in liquid nitrogen until further processing. Each specimen was sliced into 5 μm thickness section and observed by confocal microscopy. For microarray analysis, skin tissues from the mice were excised and immediately stored in liquid nitrogen in cryotubes until microarray analysis.

For each RNA sample, synthesis of the target cRNA probes and hybridization were performed using Agilent’s Low Input QuickAmp Labeling Kit (Agilent Technologies, USA) according to the manufacturer’s instructions. Briefly, each 25 ng total RNA and the T7 promoter primer were mixed and incubated at 65 °C for 10 min. The cDNA master mix (5x First strand buffer, 0.1 M DTT, 10 mM dNTP mix, RNase-Out, and MMLV-RT) was prepared and added to the reaction mixer. The samples were incubated at 40 °C for 2 h, and then dsDNA synthesis was terminated by incubating at 70 °C for 10 min. The transcription master mix was prepared as the manufacturer’s protocol (4x Transcription buffer, 0.1 M DTT, NTP mix, 50% PEG, RNase-Out, Inorganic pyrophosphatase, T7-RNA polymerase, and Cyanine 3-CTP). Transcription of the dsDNA was preformed by adding the transcription master mix to the dsDNA reaction samples and incubating at 40 °C for 2 h. Amplified and labeled cRNA was purified on an RNase mini column (Qiagen) according to the manufacturer’s protocol. The Labeled cRNA target was quantified using an ND-1000 spectrophotometer (NanoDrop Technologies, USA). After verifying labeling efficiency, each 1650 ng of cyanine 3-labeled cRNA target was carried out the fragmentation of cRNA was fragmented by adding 10x blocking agent and 25x fragmentation buffer and incubating at 60 °C for 30 min. The fragmented cRNA was resuspended with 2x hybridization buffer and directly pipetted onto the assembled Agilent Mouse (V2) Gene Expression 4 × 44 K Microarray. The arrays were hybridized at 65 °C for 17 h using hybridization oven (Agilent Technologies, USA). The hybridized microarrays were washed as indicated in the manufacturer’s washing protocol (Agilent Technologies, USA).

The hybridization images were analyzed by Agilent DNA Microarray Scanner (Agilent Technologies, USA) and data quantification was performed using Agilent Feature Extraction software 10.7 (Agilent Technologies, USA). The average fluorescence intensity of each spot was calculated and the local background was subtracted. Normalization of data and selection of fold-changed genes were performed using GeneSpringGX 7.3.1 (Agilent Technologies, USA). Normalization for Agilent one-color method was performed, which is Data transformation: Set measurements less than 5.0 to 5.0 and Per Chip: Normalize to 50th percentage. Each average normalized ratio was calculated by dividing the average of the control normalized signal intensity by the average of the test normalized signal intensity. Functional annotation of the genes was performed according to Gene Ontology™ (GO) Consortium (http://​www.​geneontology.​org/​index.​shtml) by GeneSpringGX 7.3.1.

Total RNA was extracted from mouse skin by using TRIzol reagent (Invitrogen, USA). Reverse transcription was performed on total RNA using SuperScript II reverse transcriptase (Invitrogen, USA) according to the manufacturer’s instructions. The resulting cDNA was amplified using the following primer pairs: (5′ → 3′) Tnfrsf19 (Forward: ATGACAGGGATGATCAAAGC, Reverse: TCGGCATGTGGAAAATATCT), Ercc2 (Forward: AAGAGGAGCCCAAAAAGACA, Reverse: CATCCGTGACATCAGTCAGA), Lama5 (Forward: TGCTTGAGGAAGCTGCTGAT, Reverse: CACTGCCCCCTGGATTTGTA), Ctsl (Forward: GGGACAACCACTGTGGACTT, Reverse: CTCATTACCGCTACCCATCA), Per1 (Forward: TGCATCGTCCCATTGTGAGT, Reverse: CCATGCCAGCCTGGATACTT), GAPDH (Forward: GGCATTGCTCTCAATGACAA, Reverse: ATGRAGGCCATGAGGTCCAC). Real-time PCR was performed on the StepOnePlus™ Real Time PCR System (Applied Biosystems, USA) using the SYBR Green PCR Kit (Applied Biosystems, USA), according to the manufacturer’s instructions. The thermal cycling conditions were 95 °C for 10 min followed by 40 cycles of 95 °C for 15 s and at optimal Tm (59 °C) for 30 s. The data were analyzed using the StepOne software v2.2.2 (Applied Biosystems, USA). The expression levels of each mRNAs were normalized to the endogenous control GAPDH and were calculated using the 2-ΔΔCt method.

The BisoGenet plug-in [22] from Cytoscape software version 2.7.0 [23] was used to build and visualize the networks for the top five significant genes using the respective list of significant genes of the GO categories. All available data sources in BisoGenet (including BIOGRID, DIP, BIND, and others) were selected to generate the interactions.

To investigate whether M30 could prevent CTX-induced alopecia in mice, M30 was continuously infused using subcutaneously implanted osmotic pumps in depilated C57BL/6 mice. The hair was depilated on the dorsal surface of the normal and M30-supplemented mice, and then alopecia was induced by a single intraperitoneal injection of CTX (120 mg/kg body weight). The shaved skin of the telogen mice was pink and darkened with anagen initiation. Two weeks after injection of CTX, normal black hair was apparent in the normal mice, whereas abnormal gray hair was apparent in the CTX-treated mice. However, M30 supplement prevented the growth of abnormal gray hair in the CTX-treated mice (Fig. 1a). To further study the M30 normalized abnormal gray hair; we analyzed transverse sections of the dorsal skin from 2 weeks after injection of CTX. As suggested by the microscopic observation, M30 markedly increased the depth and size of the hair follicles, prevented the dystrophic changes seen with the CTX-treated mice, and normalized the appearance of the skin to that of normal mice (Fig. 1b).

To investigate the alterations of gene expression in mouse skin during CTX and M30 treatment, we isolated total RNA from the skin of normal, CTX-treated mice, and M30 supplemented CTX-treated and we applied this RNA to the assembled Agilent Mouse (V2) Gene Expression 4 × 44 K Microarray, which contains 39,429 mouse genes. After hybridization, the microarray slide was scanned and analyzed. Each mouse gene was quantified according to its Cy3-labeled versus Cy5-labeled signal intensity. The graphs are shown (Fig. 2a) on a log scale using the Agilent Feature Extraction software 10.7 (Agilent Technologies, USA). These raw images were normalized by MA plot using locally weighted scatter plot smoothing (LOWESS) method. We applied the MA plot normalization process where M = log (Cy5/Cy3) is the log ratio of the two dyes used in the hybridization, and A = [log (Cy5) + log (Cy3)]/2 is the average of the log intensities. A skewed form of ratio pattern before normalization changed to a linear pattern centered on zero after normalization. To understand which gene expression had changed, hierarchical clustering analysis was performed by using 9373 genes as shown in Fig. 2b.

Differences in the expression patterns of the protein-coding genes in the dorsal skin among normal mice, CTX-treated mice, and M30 supplemented CTX-treated mice were analyzed. By analyzing a genome-wide microarray, we observed significant transcriptional changes in mouse skin. We compared the CTX-treated mice with normal mice (CTX/Normal), M30 supplemented CTX-treated mice with normal mice (MC/Normal), and M30 supplemented CTX-treated mice with CTX-treated mice (MC/CTX), using a filter criterion of > 1.5-fold change with p < 0.05. We then compared the results of the skin samples obtained from these groups using the hierarchical clustering method in Cluster 3.0 software. The up-regulated genes were highly expressed in the experimental group (red), and the expression levels of the down-regulated genes were significantly decreased (green) (Fig. 2b). The cluster analysis shows that CTX and M30 induce differences in gene expression in the skin of the mice, and the stylized Venn diagram depicts the patterns of changes in the gene expression levels in each skin sample. The numbers of up-regulated, contra-regulated, and down-regulated genes that responded commonly or uniquely in response to the treatment are shown with red arrows, red texture, and blue arrows, respectively (Fig. 2c).

Based on Fig. 2b, we screened out 7722 genes (3422 up-regulated and 4300 down-regulated genes) in CTX/Normal, 1148 genes (866 up-regulated and 282 down-regulated genes) in MC/Normal, and 2753 genes (1715 up-regulated and 1038 down-regulated genes) in MC/CTX, using a filter criterion at least 1.5-fold change with p < 0.05. In addition, we rescreened recovered genes by M30 treatment against the CTX-regulated genes. The M30-down-regulated 644 genes against up-regulation by CTX were selected using a filter criterion of greater than 1.5-fold change (CTX/normal) and less than 0.666-fold change (MC/CTX) with p < 0.05. Moreover, the M30-up-regulated 241 genes against down-regulation by CTX were rescreened out using a filter criterion of less than 0.666-fold change (CTX/normal) and greater than 1.5-fold change (MC/CTX) with p < 0.05 (Table 1).

In further study, the functional annotation of the genes was assessed using a GO based biological property analysis (QuickGO; https://​www.​ebi.​ac.​uk/​QuickGO/). The numbers of interesting genes were categorized as those being involved in the hair cycle, hair cycle phase, hair cycle process, hair follicle development, hair follicle maturation, hair follicle morphogenesis, and regulation of hair cycle in CTX/Normal, MC/CTX, and MC/Normal conditions, (Fig. 3a-c). Among these, the top five target genes, namely Tnfrsf19, Ercc2, Lama5, Ctsl, and Per1 were screened and the data are presented in Fig. 3d and Table 2. Tnfrsf19, Ercc2, Lama5, and Ctsl, are associated with hair cycle, hair cycle process, and hair follicle development. And Per1, Ercc2, and Ctsl, are associated with regulation of hair cycle, hair follicle maturation, and hair follicle morphogenesis, respectively (Table 2). Taken together, these results show that CTX induces the differential expression of hair cycle, and hair follicle associated genes, which were recovered by M30 treatment.

To validate the microarray results, we quantified the expression of five target genes by quantitative RT-PCR (qRT-PCR) in normal, CTX, and MC sample. All qRT-PCR analyses were performed in samples previously used for the microarray experiments. Table 3 summarize the gene expression measurements of the five validated genes by qRT-PCR. We found that both methods (microarray analysis and qRT-PCR) detected similar patterns for the five target genes by condition 1 and condition 2. The respective p-values for the microarray and qRT-PCR data were significant at the 0.05 level (Table 3).

Following combined target prediction, overlapped gene deletion and validation, five genes were identified as targeted by M30 treatment, as shown in Fig. 3d and Table 2. To gain insight into the dynamics of these five significant genes and associated genes, we mapped the gene interactions network based on datasets derived from Proteomics or Genomics experiments. The five genes and their associated 25 genes are displayed in the gene network. The CTX up-regulated genes (Tnfrsf19, Ercc2, and Lama5) were shown in red nodes and down-regulated genes (Ctsl and Per1) in blue nodes, with 25 genes in the edges, respectively (Fig. 4). The entire network were verified for interactive visualization of gene interaction networks in the Cytoscape session data. The network can be loaded and visualized using Cytoscape (refer to the Methods section).