Abstract
Cranial neural crest cells (cNCCs) comprise a multipotent population of cells that migrate into the pharyngeal arches of the vertebrate embryo and differentiate into a broad range of derivatives of the craniofacial organs. Consequently, migrating cNCCs are considered as one of the most attractive candidate sources of cells for regenerative medicine. In this study, we analyzed the gene expression profiles of cNCCs at different time points after induction by conducting three independent RNA sequencing experiments. We successfully induced cNCC formation from mouse induced pluripotent stem (miPS) cells by culturing them in neural crest inducing media for 14 days. We found that these cNCCs expressed several neural crest specifier genes but were lacking some previously reported specifiers, such as paired box 3 (Pax3), msh homeobox 1 (Msx1), and Forkhead box D3 (FoxD3), which are presumed to be essential for neural crest development in the embryo. Thus, a distinct molecular network may the control gene expression in miPS-derived cNCCs. We also found that c-Myc, ETS proto-oncogene 1, transcription factor (Ets1), and sex determining region Y-box 10 (Sox10) were only detected at 14 days after induction. Therefore, we assume that these genes would be useful markers for migratory cNCCs induced from miPS cells. Eventually, these cNCCs comprised a broad spectrum of protocadherin (Pcdh) and a disintegrin and metalloproteinase with thrombospondin motifs (Adamts) family proteins, which may be crucial in their migration.
Introduction
Stem cell-based tissue engineering is important in the field of oral science as it allows the regeneration of damaged tissues or organs [1,2]. Various stem cell populations have been identified as having a regeneration potential in the craniofacial region; however, the cranial neural crest cells (cNCCs) are considered as one of the most important candidates due to their role in craniofacial tissue organization [3].
cNCCs comprise a multipotent population of migratory cells that are unique to the vertebrate embryo and give rise to a broad range of derivatives [4,5], with the neural crest (NC) being capable of forming teratoma when transplanted into the immunocompromised animals [6]. The development of cNCCs involves three stages [7–10]: the neural plate border stage, the premigratory stage, and the migratory stage. During the migratory stage, the cNCCs delaminate from the posterior midbrain and individual rhombomeres in the hindbrain [11] and migrate into the pharyngeal arches to form skeletal elements of the face and teeth and contribute to the pharyngeal glands (thymus, thyroid, and parathyroid) [12]. Consequently, presumably cNCCs may represent a new treatment strategy for diseases in the craniofacial region [13].
Development from the premigratory to migratory stage proceeds swiftly [14], making it difficult to isolate and characterize a pure cNCC population from the embryo [15]. A recent transcriptome analysis of pure populations of sex determining region Y-box 10 (Sox10) + migratory cNCCs from chicks [16] has greatly improved our understanding of the characteristics of cNCCs, and methods for deriving NCCs from the embryonic stem (ES) cells have also been reported [17–30]; however, it remains unclear whether these cells are in the migratory stage and how long it takes to promote ES cell-derived NCCs from the pre-migratory to migratory stage.
In recent years, the use of induced pluripotent stem (iPS) cells as a revolutionary approach to the treatment of various medical conditions has gained immense attention [31,32] and iPS cells have several clear advantages over ES cells and primary cultured cNCCs as a cell source in regenerative medicine [16]. NCCs have been generated from iPS cells in numerous ways [24,33–38], with two reports having examined the differentiation of NCCs from ES or iPS cells [24,39] and two articles having described the protocol for differentiating NCCs from mouse iPS (miPS) cells [33,34]; however, few studies have investigated the changes in the properties of these NCCs overtime during the dynamic differentiation processes in the NC, in particular, during the migratory stage. Embryonic NC development depends on several environmental factors that influence the NC progenitors, regulation, and the timing of differentiation, making the elucidation of the gene regulatory network and expression profiles of miPS cell-derived cNCCs important.
Recent advances in the next-generation RNA sequencing technology (RNA-seq) have made it possible to analyze the gene expression profiles comprehensively [40–42]. Therefore, here, we used RNA-seq to investigate the gene expression landscape of cNCCs induced from miPS cells.
We treated the iPS-derived cells with cNCC induction medium for 14 days and performed triplicate RNA-seq experiments. We found that standard NC markers such as nerve growth factor receptor (Ngfr), snail family transcriptional repressor 1 (Snai1), and Snai2 were remarkably increased at 7 days after cNCC induction; whereas, the expression of the cNCC markers ETS proto-oncogene 1, transcription factor (Ets1), and Sox5, -8, -9, and -10 characteristically increased at 14 days after cNCC induction. Nestin (Nes) was upregulated throughout cNCC differentiation, as described previously [23]. In contrast, the homeobox genes such as msh homeobox 1 (Msx1), paired box 3 (Pax3), and Pax7 were not detected in the NC after a longer period of differentiation, despite their expressions having been observed in several animals [43–52]. Furthermore, the expression of Forkhead box D3 (FoxD3), which is known to be required for maintaining pluripotency in mouse ES cells [53] and is also an important NC specifier transcription factor during embryonic development, decreased over time, suggesting that it is not a cNCC specifier in iPS-derived cells.
Another important finding was the remarkable upregulation of several metzincins, including members of the disintegrin and metalloproteinase domain metallopeptidase with thrombospondin motifs (Adamts) metalloproteinase family, which play crucial roles in modulating the extracellular matrix (ECM) during development [54–56]. We assume that various kinds of Adamts proteins produce distinct extracellular proteins that are digested by cNCC swallowing them to easily migrate toward their final destinations. We also found that the expressions of nearly all procadherin (Pcdh) superfamily members were increased, some only at the migratory stage. Pcdh is the largest subfamily of cadherins and the digestion of Pcdh protein by Adam proteins is crucial for development [57].
Eventually, our results indicated that c-Myc; Ets1; Sox10; Adamts2 and -8; protocadherin alpha 2 (Pcdha2); Pcdha5, -7, -11, and -12; protocadherin alpha subfamily C,1 (Pcdhac1); and protocadherin gamma subfamily C, 3 (Pcdhgc3) may represent appropriate markers for migratory cNCCs induced from miPS cells.
Materials and Methods
miPS cell culture
All of the mouse studies were conducted in accordance with protocols approved by the Animal Research Committee of Tokyo Dental College (No. 270401).
The miPS cells that were used in this study (APS0001; iPS-MEF-Ng-20D-17 mouse induced pluripotent stem cell line) were purchased from RIKEN BRC (Ibaraki, Japan) [58]. The cells were maintained on inactivated murine embryonic fibroblast (MEF) feeder cells in Dulbecco’s Modified Eagle’s Medium (DMEM; Invitrogen, Carlsbad, CA, USA)supplemented with 15% KnockOut™ Serum Replacement (Invitrogen), 1% nonessential amino acids (Chemicon, Temecula, CA, USA), 1% L-glutamine (Chemicon), 1000 U/ml penicillin–streptomycin (P/S; Invitrogen), and 0.11 mM 2-mercaptoethanol (Wako Pure Chemical Industries Ltd., Osaka, Japan) and were passaged in 60-mm cell culture plates at a density of 1 × 105 cells/plate. The cells were grown in 5% CO2 at 95% humidity and the culture medium was changed each day.
Embryoid body (EB) formation and cNCC differentiation
We obtained cultured cNCC cells following a previously described procedure [59], as outlined in Fig 1. miPS cells were dissociated with 0.05% trypsin–ethylenediaminetetraacetic acid (EDTA; Invitrogen) and were transferred to low-attachment, 10-mm petri dishes at a density of 2 × 106 cells/plate to generate EBs. The EBs were then cultured in the NC induction medium comprising a 1:1 mixture of DMEM and F12 nutrient mixture (Invitrogen) and Neurobasal™ medium (Invitrogen) supplemented with 0.5 × N2 (Invitrogen), 0.5 × B27 (Invitrogen), 20 ng/ml basic fibroblast growth factor (Reprocell, Yokohama, Japan), 20 ng/ml epidermal growth factor (Peprotech, Offenbach, Germany), and 1% penicillin–streptomycin (P/S) for 4 days, during which time the medium was changed every other day. After 4 days, the day 0 (d0) EBs were collected and plated on 60-mm cell culture plates coated with 1μg/ml collagen type I (Advanced BioMatrix, San Diego, CA, USA). The cells were then subcultured in the same medium, which was changed every other day, and any rosetta-forming cells were eliminated. After 7–10 days, d7 cells were dissociated with 0.05% trypsin–EDTA and transferred to 60-mm cell culture plates coated with 1μg/ml collagen type I at a density of 1 × 105 cells/plate to generate 14 cells. The cells from each of these passages were collected for RNA extraction.
O9–1 cell culture
O9–1 cells, which area mouse cNCC line, were purchased from Milliopore (Billerica, MA, USA) and cultured as previously described [50] as a control.
RNA isolation and quantitative reverse transcription polymerase chain reaction analysis
The representative NC markers Ngfr, Snai1, Snai2, Sox9, and Sox10 were selected and analyzed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis. Total RNA was extracted using QIAzol® reagent (Qiagen, Valencia, CA, USA) according to the manufacturer’s protocol and RNA purity was assessed using a NanoDrop® ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), which revealed that each RNA sample had an A260/A280 ratio of >1.9. Complementary DNA (cDNA) was synthesized using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA) and qRT-PCR analysis was performed using Premix Ex Taq™ reagent (Takara Bio Inc., Otsu, Japan) according to the manufacturer’s protocol and the Applied Biosystems® 7500 Fast Real-Time PCR System, with the primer sequences presented in Table 1. All samples were normalized to levels of 18S ribosomal RNA (18S rRNA). The relative expressions of the genes of interest were analyzed using the ΔΔCt method and were compared among the groups using analysis of variance (ANOVA) followed by the Bonferroni test where the significant differences were detected among the groups. A significance level of p < 0.05 was used for all analyses and all data are expressed as means ± standard deviations (SD).
Immunohistochemistry
The cells were fixed with 4% paraformaldehyde (Wako Pure Chemical Industries Ltd.) for 15 min followed by methanol (Wako Pure Chemical Industries Ltd) for 5 min. After washing, the nonspecific binding of antibodies was blocked by adding 5% bovine serum albumin (BSA; Wako Pure Chemical Industries Ltd.) in a phosphate buffered saline (PBS) with 0.5% Triton X-100 (PBST) for 1 h. The cells were then incubated with the primary antibodies Snai1 1:50 for goat anti-rabbit (Proteintech Group, Inc. Chicago, Il, USA) and Sox10 1:500 for goat anti-mouse (Atlas Antibodies, Bromma, Sweden) in PBST for 2 nights at 4 °C. They were then incubated in the secondary antibodies fluorenscein isothiocyanate conjugated anti-rabbit IgG (Abcam, Cambridge, MA, USA) at a dilution of 1:500 for Snai1 and anti-mouse IgG (Invitrogen) at a dilution of 1:500 for Sox10 in PBST for 1 h. Eventually, the cells were stained with 4,6-diamidino-2-phenylindole (DAPI; Sigma, Livonia, MI, USA) to visualize the nuclear DNA.
RNA-seq and analysis
Total RNA from each sample was used to construct libraries with the Illumina TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA), according to the manufacturer’s instructions. Polyadenylated mRNAs are commonly extracted using oligo-dT beads, following which the RNA is often fragmented to generate reads that cover the entire length of the transcripts. The standard Illumina approach relies on randomly primed double-stranded cDNA synthesis followed by end-repair, ligation of dsDNA adapters, and PCR amplification. The multiplexed libraries were sequenced as 125-bp paired-end reads using the Illumina Hiseq2500 system (Illumina). Prior to performing any analysis, we confirmed the quality of the data and undertook read cleaning, such as adapter removal and simple quality filtering, using Trimmomatic (ver. 0.32). The paired-end reads were then mapped to the mouse genome reference sequence GRCm38 using the Burrows–Wheeler Aligner (ver. 0.7.10). The number of sequence reads that were mapped to each gene domain using SAM tools (ver. 0.1.19) was counted and the reads per kilobase of transcript per 1 million mapped reads (RPKM) for known transcripts were calculated to normalize the expression level data to gene length and library size, allowing different samples to be compared.
Results
Gene expression profiles and immunohistochemistry of cNCCs differentiated from miPS cells
The expressions of the NC markers Ngfr, Snai1, Snai2, Sox9, and Sox10 were examined by qRT-PCR in cNCCs differentiated from miPS cells as well as in O9–1 cells as a control. We detected the expression of all genes except Ngfr and Sox10 in the O9–1 cells [50]. In contrast, all five genes were detected in the cNCCs, with the premigratory neural crest markers Ngfr, Snai1, and Snai2 having the highest expression levels in d7 cells and the migratory and cranial neural crest markers Sox9 and Sox10 having the highest levels in d14 cells (Fig. 2A).
The strongest immunofluorescent staining was detected in d7 cells for Snai1 and d14 cells for Sox10 (Fig 2B).
NC specifier transcription factors
We conducted a literature search of NC specifier transcription factors that have been identified in vivo [16, 43–52, 60–106] (Tables 2 and 3) and compared these with our RNA-seq results. The relative expressions of genes that underwent a significant change in expression are presented in Fig 3A.
Open circles indicate genes that were upregulated on day 7 (d7) or d14 compared with d0 [log fold change (FC) > 1, p < 0.01, false discovery rate (FDR) < 0.05), whereas crosses indicate genes that were not upregulated.
Open circles indicate genes that were upregulated on day 7 (d7) or d14 compared with d0 [log fold change (FC) >1, p < 0.01, false discovery rate (FDR) < 0.05), whereas crosses indicate genes that were not upregulated.
We found that the transcription factor AP-2 alpha (Ap2) along with Pax3 and zinc finger protein of the cerebellum 1 (Zic1), both of which are regulated by Ap2, were most highly expressed in d7 cells (Fig 3A). Pax6, which has been reported in human ES and iPS-derived NC cells (Tables 2 and 3), was detected in both d7 and d14 cells, whereas Pax7, which has not previously been reported in the mouse NC, was also detected in the d7 cells (Fig 3A). In contrast, the homeobox genes gastrulation brain homeobox 2 (Gbx2), Msx1, distal-less homeobox 3 (Dlx3), Zic2, and Zic3 were not detected in d7 or d14 cells, and the homeobox genes Zic1 and Dlx5 were only expressed in d7 cells, despite these having been reported in the NC of a range of species (Table 2); however, Meis homeobox 2 (Meis2) was expressed in both d7 and d14 cells.
Both MYCN proto-oncogene, bHLH transcription factor (N-myc) and c-Myc have been reported in NCCs (Table 3); however, we did not observe N-myc expression in d7 or d14 cells and detected c-Myc expression in the d7 and d14 group (Fig 3A). Furthermore, we observed substantial downregulation of the winged-helix transcription factor FoxD3 over time (Fig 3A), which is an important factor for maintaining the pluripotency of ES cells and a key NC specifier that has been implicated in multiple steps of NC development and NCC migration in the embryo of various species (Table 2).
The premigratory NC markers Ngfr, heart and neural crest derivatives expressed 2 (Hand2), Snai1, and Snai2 were only detected in the d7 cells; however, other premigratory NC markers, such as platelet derived growth factor receptor, alpha polypeptide (Pdgfra), 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 (Pfkfb4), inhibitor of DNA binding 2 (Id2), Id3, and Id4 were found in both d7 and d14 cells, as was Nes (Fig 3A).
Migratory neural crest markers expression of Sox5, -6, -8, -9, and -10, which encode members of the sex-determining region Y (SRY)-related high mobility group (HMG)-box family of transcription factors and have been reported to be crucial in several aspects of NCCs, were observed in d7 or d14 cells. Sox10, which is a known marker of migratory cNCCs in various species (Table 2), was only detected in d14 cells, as were other migratory NC markers, including retinoid X receptor gamma (Rxrg), collagen type IX alpha 3 chain (Col9a3), and endothelin receptor type B (Ednrb; Fig. 3A); however, LIM domain only 4 (Lmo4), leukocyte tyrosine kinase (Ltk), erb-b2 receptor tyrosine kinase 3 (Erbb3), and angiogenin, ribonuclease A family, member 2 (Ang2) were not detected in the d14 cells.
Twist family bHLH transcription factor 1 (Twist1), which is activated by a variety of signal transduction pathways and is crucial in the downregulation of E-cadherin expression, was detected in both d7 and d14 cells, as was beta-1,3-glucuronyltransferase 1 (B3gat1/Hnk1), or CD57. In contrast, expression of the trunk NC markers lit guidance ligand 1/2 (Slit1/2), which has been reported to play an important role in the migration of trunk NC cells toward ventral sites, was upregulated only in the d7 cells (Fig 3A).
Eventually, the expressions of tenascin C (Tnc), cadherin-6 (Cdh6), and ras homolog family member B (Rhob), all of which are related to cell adhesion and motility [106–111], significantly increased in both d7 and d14 cells (Fig 3B).
Metzincin superfamily zinc proteinase and protocadherin superfamily expressions
Members of the metzincin superfamily are proteinases that have a zinc ion at their active site. This family includes the matrix metalloproteinases (Mmps), a disintegrin and metalloproteinase (Adam), and Adamts, all of which have attracted attention as factors involved in cancer cell invasion and cell migration. Mmp2, -11, -14, -15, -16, -24, and -28 were significantly upregulated in the cNNCs (Fig 4A), all of which except Mmp24 are membrane-bound types. The expressions of Mmp11 and -28 were only detected in d7 cells, while all other Mmps were detected in both d7 and d14 cells (Fig 4A, B).
Only Adam1a, -8, -10, and -12 were upregulated in both d7 and d14 cells (Fig 4C, D), despite the members of this family being important in NC migration and the expressions of Adam10, -12, -15, -19, and-33 having been observed in the mouse NC [112]. In contrast, various Adamts family genes, which are important for connective tissue organization and cell migration, were upregulated in either d7 or d14 cells (Fig 4C, D). The expression of Adamts1 in particular exhibited a substantial increase in expression, while Adamts2 and -8, which are presumed to be important in cancer cell invasion [55], increased in the later stages of differentiation.
The vertical axis reveals reads per kilobase of exon per million mapped reads (RPKM) and the horizontal axis indicates time. Each experiment was performed in triplicate with values representing the mean ± SD. Groups were compared using ANOVA followed by the Bonferroni test: *p < 0.05. (C) Expressions of Adam and Adamts genes in mouse. Round marks alongside d7 or d14 indicate that the genes were upregulated compared with d0 (logFC > 1, p < 0.01, FDR < 0.05), whereas cross marks indicate no upregulation. (D) Graphical representation of the upregulation of Adam1a and 8–12, and Adamts1–10, -12, and 15–20 in d7 or d14 cells. Adam2, -4, -7, and -8, and Adamts 9 and -12 were most upregulated in d14 cells. The vertical axis reveals reads per kilobase of exon per million mapped reads (RPKM) and the horizontal axis indicates time. Each experiment was performed in triplicate with values representing the mean ± SD. Groups were compared using ANOVA followed by the Bonferroni test: *p < 0.05.
Most of the Pcdh genes, which are involved in cell adhesion, were upregulated in d7 and d14 cells (Table 4); however, Pcdha2, -5, -7, -11, and -12; Pcdhac1; and Pcdhgc5 were only upregulated in the d14 cells.
Discussion
In this study, we successfully generated miPS-induced cNCCs that were sufficiently close to the migratory stage. The NC has previously been generated from ES or iPS cells in various ways [24,33–39] and the protocol we used in the present study was based on the methods outlined by R. Bajpai et al. [39]; however, few studies have investigated the changes in the properties of cNCCs at different time points (Table 5).
Our d7 and d14 cells expressed typical NC markers, such as Ngfr, Snai1, and Snai2. In contrast, the mouse cNCC line (O9–1 cells) did not express Ngfr, indicating that cNCCs derived from miPS cells may be of better quality for evaluating the cNCC characteristics than O9–1 cells [59]. We also found that, unlike O9–1 cells, d14 cells expressed considerably high levels of Sox10, which is considered as a reliable marker for migratory cNCCs. Because cNCCs are involved in organizing numerous craniofacial tissues, several reports are available on their gene expression profiles; however, we found that various results that have been reported were inconsistent between the species and protocols. Since cNCCs differentiate fast in the embryo [14], it is considerably difficult to synchronize the timing of isolation to a particular point in their development. Furthermore, migratory cNCCs intermingle with other types of cells in the embryo, making it difficult to isolate and characterize a pure cell population. Consequently, there have been few reports of cNCC markers [16,60–71]; however, Simoes-Costa et al. [16] successfully isolated Sox10 positive cNCCs in a chicken embryo and analyzed their gene profiles and we found that d14 cells expressed several of these Sox10 positive chicken cNCCs. It has previously been suggested that NC cells have multiple populations [11] and that the generation of cNCCs from iPS cells could result in numerous different populations occurring in the same dish. Therefore, this diversity in populations may explain the discrepancies; however, we can conclude that under the conditions used in the present study, cMyc; Ets1; Sox10; Adamts2; Adamts8; Pcdha2, -5, -7, -11, and-12; Pcdhac1, and Pcdhgc3 may represent useful markers for migratory cNCCs.
Our results also indicated that d7 cells were still in the premigratory stage even though they expressed numerous NC markers. Thus, cNCCs derived from miPS cells took more than 14 days to become migratory in vitro, which is much slower than has been observed in the mouse embryos in vivo under the same conditions [113].
RNA-seq makes it possible to normalize the expression levels of different genes, allowing comparisons between samples. We conducted triplicate experiments in which none of the induced cNCCs expressed several homeobox genes that are considered to be expressed in the early stages of cNCC differentiation. In particular, we did not observe FoxD3 expression in either d7 or d14 cells, despite it being recognized as one of the key transcription factors in cNCCs [53]. These negative results indicate that cNCCs derived from miPS cells may have distinct gene regulatory networks. Although it is possible that the cells would express those genes at different time points, the expression of FoxD3, which is a pluripotent stem cell marker gene and plays an important role in maintaining pluripotency, decreases in a time-dependent manner [44], making it more likely that FoxD3 may not be a key regulator in iPS-derived cNCCs. We speculate, however, that iPS cells had a sufficient amount of FoxD3 to allow them to be converted from iPS cells into cNCCs.
Protocadherins belong to the cadherin superfamily and are involved in intercellular interactions [57], while metzincins are thought to be key proteinases that facilitate the cell migration [45]. Unfortunately, the abundances of members of these families hindered their analysis; however, since RNA-seq techniques enable us to evaluate the gene profiles exhaustively, we were able to focus on the expressions of all of the procadherin and metazicin family members. As expected, we found that several Adam and Adamts genes were upregulated, with most of the latter increasing significantly. The Adam genes that increased in the cNCCs were the membrane-bound type; whereas, the Adamts genes were secreted proteinases, indicating that the expression of various Adamts may allow the matrix to be digested more efficiently, as each may be capable of digesting a different type of extracellular matrix protein [45]. Thus, the secretion of a variety of Adamts and Pcdh proteins may play a crucial role in the migration ability of cNCCs.
In summary, we successfully induced the formation of cNCCs from miPS cells by placing them in NC inducing media for 14 days. We found that although the resulting cNCCs had several NC specifiers, some were lacking, indicating that a distinct molecular network may control the gene expression in miPS-derived cNCCs. Our results also indicated that cMyc; Ets1; Sox10; Adamts2 and -8; Pcdha2, -5, -7, -11, and -12; Pcdhac1; and Pcdhgc3 may represent appropriate markers for migratory cNCCs induced from miPS cells. Eventually, these cNCCs produced a broad spectrum of Adamts family proteins that may play an important role in their migration.
Conflict of interest
The authors have no conflicts of interest directly relevant to the content of this article.
Acknowledgments
The author is grateful to Professor T. Azuma, MD, PhD, Department of Biochemistry, and Professor T. Ichinohe, DDS, PhD, Department of Dental Anesthesiology, for their guidance. I also thank S. Onodera and A. Saito, Department of Biochemistry.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.
- 9.
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.↵
- 24.↵
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.
- 36.
- 37.
- 38.↵
- 39.↵
- 40.↵
- 41.
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.
- 47.
- 48.
- 49.
- 50.↵
- 51.
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.
- 62.
- 63.
- 64.
- 65.
- 66.
- 67.
- 68.
- 69.
- 70.
- 71.↵
- 72.
- 73.
- 74.
- 75.
- 76.
- 77.
- 78.
- 79.
- 80.
- 81.
- 82.
- 83.
- 84.
- 85.
- 86.
- 87.
- 88.
- 89.
- 90.
- 91.
- 92.
- 93.
- 94.
- 95.
- 96.
- 97.
- 98.
- 99.
- 100.
- 101.
- 102.
- 103.
- 104.
- 105.
- 106.↵
- 107.
- 108.
- 109.
- 110.
- 111.↵
- 112.↵
- 113.↵