ORIGINAL ARTICLES
Immunohistochemical Study on the Developing Optic Nerves in Human Embryos and Fetuses Shozo Takayama, MD, Misao Yamamoto, MD, Kimio Hashimoto, MD and Hiroshi Itoh, MD
For the purpose of investigating the time of development and myelination of the optic nerve, we applied immunohistochemical stains to the optic nerves of 26 human embryos and fetuses aged 6-39 weeks. The minimum diameter of the optic nerve's cross section increased with the development of embryos; 0.15 mm at the 9th week, 0.4 mm at the 11th week, 1.5 mm at the 24th week, and 2.1 mm at the 32nd week. With regard to myelin basic protein (MBP), the localization could not be observed at 28 gestational weeks. The MBP was proved to be positive in oligodendrocytes of the optic nerve at the age of the 32nd gestational week. The localization of glial fibrillary acidic protein (GFAP) was proved partially in astrocytes around the optic nerve in the case of 24 gestational weeks. The remarkable localization ofGFAPpositive astrocytes inside the nerve fibers was proved at the 27th gestational week. The histological appearance ofGFAPpositive cells advanced to that of MBP-positive ones after 8 weeks interval. The degree of MBP staining was less marked than that of GFAP in examined fetal cases. These results suggest that astrocytes and oligodendrocytes correlate to the developmental growth and myelination of the optic nerves during the fetal development. Key words: Immuno-histochemical stains, myelination, myelin basic protein (MBP), glilll fibrillary acidic protein (GFAP), human optic nerves. Takayama S, Yamamoto M, Hashimoto K, Itoh H. Immunohistochemical study on the developing optic nerves in human embryos and fetuses. Brain Dev 1991;13:307-12
It is considered that astrocytes and oligodendrocytes may differentiate from neuroepithelial cells, and that they both participate in the development and myelination of optic nerve cells. However, there are no established theories about when these cells begin to appear. Based on observations from 3 human optic nerves with an electron microscope, Cravioto [1] reported that the optic nerves were composed of large bundles ofaxons surrounded by parallel rows of glial cells at the 12th gestational week, and that their myelination began apparently between the 14ih and 22nd gestational week.
From the Departments of Ophthalmology (ST) and Pathology (KH), Kobe Children's Hospital, Kobe; Departments of Ophthalmology (MY) and Pathology (HI), Kobe University School of Medicine, Kobe. Received for publication: March 11, 1991. Accepted for publication: July 9, 1991. Correspondence address: Dr. Shozo Takayama, Department of Ophthalmology, Kobe Children's Hospital, 1-1-1 Takakuradai, Suma-ku, Kobe 654, Japan.
To clarify the developmental background of the optic nerve relative to glial cells and myelination, 26 human embryos aged 6-39 weeks were examined pathologically using immunohistochemical stains. MATERIALS AND METHODS Materials The examined cases were the pre chiasmatic nerves of 9 autopsied fetuses (24-39 gestational weeks) and 17 aborted embryos and fetuses (6-21 gestational weeks) during the period from September 1982 to April 1984. Since the latter 17 cases were the products of unknown or incorrectly estimated periods of pregnancy, we calculated the age according to the conversion table based on Iwamoto's crown-rump length (CRL) [2]. Two autopsied cases, a one-month-old baby and a fifteen-year-old adolescent, were selected as the control subjects. For the pathological examination of optic nerves, we prepared cross sections of about 1-3 mm to the eye ball and vertical ones immediately to the center. In some
cases, vertical specimens with the eye ball inducted were prepared. On the examination of cross sections, their minimum diameters were measured respectively using their histological sections. Methods After having fIxed the specimens with 10% buffered formalin, we dehydrated them with graded alcohol and then embedded tl).em in paraffIn. Serial sections of 3-411m thickness were prepared. Immunohistochemical stains of myelin basic protein (MBP) and glial fibrillary acidic protein (GFAP) by the peroxidase anti-peroxidase (PAP) procedure were used for these specimens_ The antibodies used were anti-MBP antibody (polyclonal, Lipshow Detroit) and anti-GFAP antibody (monoclonal, Lipshow Detroit). The procedure for the irnmunostaining was as follows: (1) After having removed the paraffIn from the specimen which was prepared through formalin fixation, it was soaked in a 3% H202 solution for five minutes to block the endogenous peroxidase within tissues . Then it was rinsed in phosphate-buffered saline (PBS). (2) Normal goat serum was applied to it for twenty minutes. (3) Then rabbit anti-MBP serum and anti-GF AP serum, which were the primary antibodies, were applied for twenty minutes. Then, it was rinsed in PBS. (4) It was reacted with goat anti-rabbit serum, which was the linking reagent, for twenty minutes. Then it was rinsed in PBS. (5) It was applied to rabbit PAP complex for twenty minutes. Then it was rinsed in PBS. (6) It was applied to 0.005% H202 and 3.3' diaminobenzidine (DAB) solution (Karnovsky solution) of pH 7.6 as substrate solution for fIve minutes. After confIrming the appearance of dark-brown reaction under. the microscope, the reaction was stopped by PBS. We then counter-stained with hematoxylin, dehydrated with alcohol, flooded with xylol, and mounted on glass slides using permanent aqueous medium and balsam. RESULTS The staining results with MBP and GF AP for each case are shown in Table 1. In optic nerves of 17 embryos and fetuses aged 6 to 21 weeks, there was no staining with MBP or GFAP. At the 9th gestational week, the minimum diameter of cross section of the optic nerve was 0.15 mm. It could be observed in case No 8 that the axons which grew from the ganglion cells in the retina were forming the nerve fIber layer. However, the presence of MBP and GF AP could not be proved in these nerve fIbers. MBP and GF AP in the optic nerve at this time were both negative. The nerve was loosely subdivided by thin fIbrovascular cords without identifiable oligodendrocytes and microglial cells. At the 11 th gestational week, the minimum diameter of the cross-sectioned optic nerve was 0.4 mm. It was confIrmed in case No 10 that there was a tendency in the
308 Brain & Development, Vol 13, No 5, 1991
nerve fibers to divide into small bundles of fIbers. MBP and GF AP stainings were both negative. At the 15th gestational week, the diameter of the optic nerve reached to 1.0 mm. Fig 1 is a cross section of the optic nerve in case No. 14. MBP and GFAP were also negative, and there were no evident oligodendrocytes or microglial cells. At the 24th gestational week, the diameter of the optic nerve was 1.5 mrn. Fig 2 is the cross section of the optic nerve in case No 18. MBP was negative but GFAP was positive in a few nerve fIbers. The nerve stroma was edematous but a few oligodendrocytes and microglial cells which might be induced by secondary degeneration of the optic nerve directly connected to their circulatory disturbance before their death, were present. At the 32nd gestational week, the minimum diameter of the cross-sectioned optic nerve was about 2.1 mm. Fig 3 is obtained from case No 24. Staining for both GFAP and MBP was positive. It was confIrmed that there were cells which could be regarded as oligodendrocytes possessing a small and round nucleus with perinuclear halo. At the same time, nerve fibers were close together
Table I Summary of cases Case
2 3 4
5 6 7 8 9 10
Gestational weeks
11
13
15 15 15 19 20 21 24 24 27 27 27 28 32 32 39
13
GFAP stain
6 6 7 8 8 8 8 9 9 11
12 14 15 16 17 18 19 20 21 22 23 24 25 26
MBP stain
+ + +
+ + + + + + + + +
Diagnosis
Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Hydrocephalus Pulmonary abscess RDS RDS Skeletal anomalies Sepsis Sepsis Prematurity Sepsis
and had become thicker. Many more GFAP-positive cells were recognized along the axonal bundles. Toward the last stage of fetal life, the optic nerve
became thicker and it was about 2.7 mm at about the 39th week of fetal life. Fig 4A is a longitudinal section of the optic nerve in case No 26. MBP-positive cells increas-
Fig 1 Cross section of the optic nerve at 15 gestational weeks, negative for MBP (A) and GFAP (B) stains_ x 200.
Fig 2 Cross section of the optic nerve at 24 gestational weeks, negative for MBP (A) but positive for GFAP (B) stains, micro glias (arrow heads), astrocytes (thin arrows). x 200.
Takayama et al: Immunohistochemistry of optic nerve 309
ed gradually in number and the nerve fibers became more diffusely stained . Fig 4B shows a longitudinal section of the optic nerve in one case of control subjects, a one-
month-old baby. MBP-positive cells decreased considerably in number and well matured myelinated fibers were observed.
Fig 3 Cross section of the optic nerve at 32 gestational weeks, positive for MBP (A) and GFAP (B) stains, astrocyte (thin arrow) . x 200.
Fig 4 Longitudinal section of the optic nerve at 39 gestational weeks (A) and in one case of control subjects, a onemonth-old baby (B), positive for MBP stain , oligodendrocytes (thick arrows). x 200.
310 Brain & Development, Vol 13, No 5, 1991
DISCUSSION There are a number of reports on the myelination of the optic nerves [1, 3-19]. According to Duke-Elder and Cook [3], there was no evidence of the process which depended on oligodendrocytes until about the 20th gestational week when meylination was evident in the geniculate bodies and some time during the 24th gestational week it was recognizable in the tract. During the 24th-28th gestational week myelination reached the chiasma, in 32 weeks it was usually apparent in the optic nerve, and it slowly extended distally to reach the level of the lamina cribrosa at birth or some weeks thereafter. Also in the reports by Magoon and Robb [4] and Nakayama [5,6], the myelination began in the 28th-32nd gestational week. On the other hand, there were reports by Bernheimer [7] and Sattler [8] that they reported myelination in the optic tracts but not optic nerves of fetuses aged 28-32 gestational weeks under light microscope. As a result of our present investigation, like the former reports, MBP-positive cells were found in the optic nerves of the fetus at 32 weeks. This result indicates that the myelination began about the 32nd gestational week. It was supported by the fact that we examined 23 embryos and fetuses of 6-28 weeks and MBP was negative in all the cases examined. With regard to the time of appearance of glial cells, Nakaizumi [9] examined 60 cases of human fetuses aged up to 28 weeks histologically. He stated that the glial cells developed from the "inner layer" of the optic stalks after 12 gestational weeks. Raff et al [20] examined cell suspensions and cultures of optic nerves of rats and summarized that primitive glial cells develop as two distinct lineages which diverge to form astrocytes and oligodendrocytes prenatally. David et al [21] provided evidence that the rat optic nerve contains three types of macroglial cells that develop as two distinct lineages: one
(mm) 3
< Growth of Optic Nerve>
gestational weeks
Fig 5 There is a rapid increase in the size of optic nerve after 24 gestational weeks.
lineage comprises type 1 astrocytes, which develop before birth, while the other comprises oligodendrocytes and type 2 astrocytes, which develop after birth from a common, bipotential glial progenitor cell. GF AP-potitive cells were also observed in our present investigation in a case at 24 gestational weeks. This means that around this stage, as David et al [21] reported, one lineage comprises type 1 astrocytes and also, as Kuwabara [10] stated, the neuroepithelial cells of the fissure portion of the optic stalks are eliminated by marked bionecrosis. It is to be considered that it is at this time that a small number of the surviving cells differentiate into type 1 astrocytes, and that they divide the nerve fibers into small axon bundles and participate in the development of the optic nerves by surrounding them. In fact, during the period after 24 weeks till the time when MBP positive cells appear, the optic nerves increase their territory at a speed of twice as fast as before. Also after the appearance of MBP-positive cells, the rate of development increases twice as fast as before (Fig 5). Following the increase of the cross sectional area of the optic nerve, MBP-positive cells and GF AP-positive cells appear, and gradually the numbers of these cells and the positivity of stained nerve fibers increases. This fact indicates that these cells participate in the full growth of the optic nerves. Recently, Choi et al [22-24] suggested that radial glias may give rise to both astrocyte and oligodendrocyte with correlative use of immunohistochemical methods, electron microscopy and Golgi methods in the human fetal spinal cord at 6 to 17 weeks as well as in human fetal cerebrum and cerebellum. They showed that "transitional forms" between astrocyte and oligodendrocyte may exist. These "transitional forms" cells had mutually exclusive identifying features of astrocytes and oligodendrocytes respectively under electron microscope and demonstrated reverse immune reaction for MBP or GFAP on the contrary to their morphological features. In this report, further discussion to the peripheral or central developing theories of the optic nerve is out of the main problem since precise examination to the whole optic nerve is not done here. However, if radial glia or radial glial cell-derived astrocyte actually undergo differentiation or transformation into myelin-forming oligodendrocyte after the release of some still undetermined Signal in accordance with the "critical diameter" of axon as Choi et al reported, it can be suggested that the myelination does not necessarily occur from the center to the periphery. Their hypothesis may be supported by the fact that MBP-positive cells increased in number and crowded here and there, where nerve fibers became stained diffusely to be seen in the optic nerve of case No 26. As a result of our present study, MBP and GFAP immunostains were both negative until 21 gestational weeks, and the differentiation into astrocytes could not be ob-
Takayama et al: Immunohistochemistry of optic nerve 311
served. As reported by Silver and others [25] it was considered that this was the stage when axons invaded the intercellular spaces and created marked bionecrosis of many neuroepithelial cells in the fissure portion of the optic stalk. Probably before this stage, the morphological and functional growth of the nerve fibers was not achieved. Among the studied embryos younger than 24 weeks, some cases with strong tissue dissolution were included, and there is the possibility that the characteristics of antigen were lost due to this and left no stains. Moreover, there was a report by Cravioto [1] that he observed the optic nerves in 14-22 gestational weeks using an electron microscope, and that he confirmed the obvious myelination. Taking these factors into consideration, a careful selection of cases and an investigation in detail should be our future research. ACKNOWLEDGMENTS The authors wish to express their sincere gratitude to Dr. J. Takenouchi, Meimai Central Hospital, and Dr. K. Fukutome, Wakamiya Hospital for their kind cooperation for this research. REFERENCES 1. Cravioto H. Electron microscopic studies on the developing human nervous system. II. the optic nerves. J Neuropathol E'xpNeuroI1965;24:164-7. 2. Iwamoto K. Mean weeks of fetal life according to each calculation of CRL (in Japanese). Nihon Sanka Fujinka Gakkai Zasshi (Tokyo) 1983;35:2330. 3. Duke-Elder S, Cook C. System of ophthalmology. Vol III. London: Henry Kimpton, 1963: 109. 4. Magoon EH, Robb RM. Development of meylin in human optic nerve and tract: a light and electron microscopic study. Arch aphthalmol 1981;99:655-9. 5. Nakayama K. Studies of the myelination of the human optic nerve. Jpn J aphthalmol (Tokyo) 1968; 11 :132-40. 6. Nakayama K. Studies on the human optic nerve, especially its myelinization (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1967;70:1511-25. 7. Bernheimer S. Uber die Entwicklung und den Verlauf der Markfasern in Chiasm nervorum opticorum des Menschen. Arch Augenheilk 1889;20:133-79. 8. Sattler CH. Uber die Markscheidenentwicklung im Tractus opticus, Chiasm und Nervus opticus. Graefes Arch Qin Exp 1915;90:271-98.
312 Brain & Development, Vol 13, No 5, 1991
9. Nakaizurni Y. Embryological study of optic nerve of the Japanese embryo (in Japanese). Rinsho Ganka (Tokyo) 1958; 12:1329-42. 10. Kuwabara T. Development of the optic nerve of the rat. Invest aphthalmol Vis Sci 1975; 14:732-45. 11. Yamamoto T. Electron-microscopic observation of human optic nerves (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1965;69:1527-80. 12. Kanazawa S. Electron microscopic study of the human fetal optic nerve (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1970; 73 :1330-53. 13. Naumann GOH. Pathologie des Auges. Berlin: SpringerVerlag, 1980. 14. Friede RL, Hu KH. Proximo-distal differences in myelin development in human optic fibers. Z Zellforschung 1967; 79:259-64. 15. Kennedy PGE, Lisak RP, Raff MC. Cen type-specific markers for human glial and neuronal cells in culture. Lab Invest 1980;43:342-51. 16. Quitschke W, Sibony P, Schechter N. Variable expressio'n of intermediate filament proteins during embryonic development of human optic nerve. Exp Neurol 1985;90:204-14. 17. Okuda T. The development of myelin formation in the human optic nerve (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1985;89:1156-65. 18. Shirai S. Embryology of optic nerve (in Japanese). Shinkei Ganka (Kanagawa) 1988;5:244-50. 19. Peters A. Observation on the connections between myelin sheaths and glial cells in the optic nerve of young rats. J Anat 1964;98: 125-34. 20. Raff MC, Abney ER, Miller RH. Two glial cell lineages diverge prenatally in rat optic nerve. Dev Bioi 1984;106: 53-60. 21. David S, Miller RH, Patel R, Raff MC. Effects of neonatal transection on glial cell development in the rat optic nerve: evidence that the oligodendrocyte-type 2 astrocyte cell lineage depends on axons for its survival. J Neurocytol 1984; 13:961-74. 22. Choi BH, Kim RC. Expression of glial fibrillary acidic protein in immature oligodendroglia. Science 1984;223:407-8. 23. Choi BH, Kim RC, Lapham LW. Do radial glia give rise to both astroglia and oligodendroglial cells? Dev Brain Res 1983;8:119-30. 24. Choi BH. Radial glia of developing human fetal spinal cord: Golgi, immunohistochemical and electron microscopic study. DevBrainRes 1981;1:249-67. 25. Silver J, Robb RM. Studies on the development of the eye cup and optic nerve in normal mice and in mutants with congenital optic nerve aplasia. Dev Bioi 1979 ;68: 175-90.
Immunohistochemical Study on the Developing Optic Nerves in Human Embryos and Fetuses Shozo Takayama, MD, Misao Yamamoto, MD, Kimio Hashimoto, MD and Hiroshi Itoh, MD
For the purpose of investigating the time of development and myelination of the optic nerve, we applied immunohistochemical stains to the optic nerves of 26 human embryos and fetuses aged 6-39 weeks. The minimum diameter of the optic nerve's cross section increased with the development of embryos; 0.15 mm at the 9th week, 0.4 mm at the 11th week, 1.5 mm at the 24th week, and 2.1 mm at the 32nd week. With regard to myelin basic protein (MBP), the localization could not be observed at 28 gestational weeks. The MBP was proved to be positive in oligodendrocytes of the optic nerve at the age of the 32nd gestational week. The localization of glial fibrillary acidic protein (GFAP) was proved partially in astrocytes around the optic nerve in the case of 24 gestational weeks. The remarkable localization ofGFAPpositive astrocytes inside the nerve fibers was proved at the 27th gestational week. The histological appearance ofGFAPpositive cells advanced to that of MBP-positive ones after 8 weeks interval. The degree of MBP staining was less marked than that of GFAP in examined fetal cases. These results suggest that astrocytes and oligodendrocytes correlate to the developmental growth and myelination of the optic nerves during the fetal development. Key words: Immuno-histochemical stains, myelination, myelin basic protein (MBP), glilll fibrillary acidic protein (GFAP), human optic nerves. Takayama S, Yamamoto M, Hashimoto K, Itoh H. Immunohistochemical study on the developing optic nerves in human embryos and fetuses. Brain Dev 1991;13:307-12
It is considered that astrocytes and oligodendrocytes may differentiate from neuroepithelial cells, and that they both participate in the development and myelination of optic nerve cells. However, there are no established theories about when these cells begin to appear. Based on observations from 3 human optic nerves with an electron microscope, Cravioto [1] reported that the optic nerves were composed of large bundles ofaxons surrounded by parallel rows of glial cells at the 12th gestational week, and that their myelination began apparently between the 14ih and 22nd gestational week.
From the Departments of Ophthalmology (ST) and Pathology (KH), Kobe Children's Hospital, Kobe; Departments of Ophthalmology (MY) and Pathology (HI), Kobe University School of Medicine, Kobe. Received for publication: March 11, 1991. Accepted for publication: July 9, 1991. Correspondence address: Dr. Shozo Takayama, Department of Ophthalmology, Kobe Children's Hospital, 1-1-1 Takakuradai, Suma-ku, Kobe 654, Japan.
To clarify the developmental background of the optic nerve relative to glial cells and myelination, 26 human embryos aged 6-39 weeks were examined pathologically using immunohistochemical stains. MATERIALS AND METHODS Materials The examined cases were the pre chiasmatic nerves of 9 autopsied fetuses (24-39 gestational weeks) and 17 aborted embryos and fetuses (6-21 gestational weeks) during the period from September 1982 to April 1984. Since the latter 17 cases were the products of unknown or incorrectly estimated periods of pregnancy, we calculated the age according to the conversion table based on Iwamoto's crown-rump length (CRL) [2]. Two autopsied cases, a one-month-old baby and a fifteen-year-old adolescent, were selected as the control subjects. For the pathological examination of optic nerves, we prepared cross sections of about 1-3 mm to the eye ball and vertical ones immediately to the center. In some
cases, vertical specimens with the eye ball inducted were prepared. On the examination of cross sections, their minimum diameters were measured respectively using their histological sections. Methods After having fIxed the specimens with 10% buffered formalin, we dehydrated them with graded alcohol and then embedded tl).em in paraffIn. Serial sections of 3-411m thickness were prepared. Immunohistochemical stains of myelin basic protein (MBP) and glial fibrillary acidic protein (GFAP) by the peroxidase anti-peroxidase (PAP) procedure were used for these specimens_ The antibodies used were anti-MBP antibody (polyclonal, Lipshow Detroit) and anti-GFAP antibody (monoclonal, Lipshow Detroit). The procedure for the irnmunostaining was as follows: (1) After having removed the paraffIn from the specimen which was prepared through formalin fixation, it was soaked in a 3% H202 solution for five minutes to block the endogenous peroxidase within tissues . Then it was rinsed in phosphate-buffered saline (PBS). (2) Normal goat serum was applied to it for twenty minutes. (3) Then rabbit anti-MBP serum and anti-GF AP serum, which were the primary antibodies, were applied for twenty minutes. Then, it was rinsed in PBS. (4) It was reacted with goat anti-rabbit serum, which was the linking reagent, for twenty minutes. Then it was rinsed in PBS. (5) It was applied to rabbit PAP complex for twenty minutes. Then it was rinsed in PBS. (6) It was applied to 0.005% H202 and 3.3' diaminobenzidine (DAB) solution (Karnovsky solution) of pH 7.6 as substrate solution for fIve minutes. After confIrming the appearance of dark-brown reaction under. the microscope, the reaction was stopped by PBS. We then counter-stained with hematoxylin, dehydrated with alcohol, flooded with xylol, and mounted on glass slides using permanent aqueous medium and balsam. RESULTS The staining results with MBP and GF AP for each case are shown in Table 1. In optic nerves of 17 embryos and fetuses aged 6 to 21 weeks, there was no staining with MBP or GFAP. At the 9th gestational week, the minimum diameter of cross section of the optic nerve was 0.15 mm. It could be observed in case No 8 that the axons which grew from the ganglion cells in the retina were forming the nerve fIber layer. However, the presence of MBP and GF AP could not be proved in these nerve fIbers. MBP and GF AP in the optic nerve at this time were both negative. The nerve was loosely subdivided by thin fIbrovascular cords without identifiable oligodendrocytes and microglial cells. At the 11 th gestational week, the minimum diameter of the cross-sectioned optic nerve was 0.4 mm. It was confIrmed in case No 10 that there was a tendency in the
308 Brain & Development, Vol 13, No 5, 1991
nerve fibers to divide into small bundles of fIbers. MBP and GF AP stainings were both negative. At the 15th gestational week, the diameter of the optic nerve reached to 1.0 mm. Fig 1 is a cross section of the optic nerve in case No. 14. MBP and GFAP were also negative, and there were no evident oligodendrocytes or microglial cells. At the 24th gestational week, the diameter of the optic nerve was 1.5 mrn. Fig 2 is the cross section of the optic nerve in case No 18. MBP was negative but GFAP was positive in a few nerve fIbers. The nerve stroma was edematous but a few oligodendrocytes and microglial cells which might be induced by secondary degeneration of the optic nerve directly connected to their circulatory disturbance before their death, were present. At the 32nd gestational week, the minimum diameter of the cross-sectioned optic nerve was about 2.1 mm. Fig 3 is obtained from case No 24. Staining for both GFAP and MBP was positive. It was confIrmed that there were cells which could be regarded as oligodendrocytes possessing a small and round nucleus with perinuclear halo. At the same time, nerve fibers were close together
Table I Summary of cases Case
2 3 4
5 6 7 8 9 10
Gestational weeks
11
13
15 15 15 19 20 21 24 24 27 27 27 28 32 32 39
13
GFAP stain
6 6 7 8 8 8 8 9 9 11
12 14 15 16 17 18 19 20 21 22 23 24 25 26
MBP stain
+ + +
+ + + + + + + + +
Diagnosis
Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Abortion Hydrocephalus Pulmonary abscess RDS RDS Skeletal anomalies Sepsis Sepsis Prematurity Sepsis
and had become thicker. Many more GFAP-positive cells were recognized along the axonal bundles. Toward the last stage of fetal life, the optic nerve
became thicker and it was about 2.7 mm at about the 39th week of fetal life. Fig 4A is a longitudinal section of the optic nerve in case No 26. MBP-positive cells increas-
Fig 1 Cross section of the optic nerve at 15 gestational weeks, negative for MBP (A) and GFAP (B) stains_ x 200.
Fig 2 Cross section of the optic nerve at 24 gestational weeks, negative for MBP (A) but positive for GFAP (B) stains, micro glias (arrow heads), astrocytes (thin arrows). x 200.
Takayama et al: Immunohistochemistry of optic nerve 309
ed gradually in number and the nerve fibers became more diffusely stained . Fig 4B shows a longitudinal section of the optic nerve in one case of control subjects, a one-
month-old baby. MBP-positive cells decreased considerably in number and well matured myelinated fibers were observed.
Fig 3 Cross section of the optic nerve at 32 gestational weeks, positive for MBP (A) and GFAP (B) stains, astrocyte (thin arrow) . x 200.
Fig 4 Longitudinal section of the optic nerve at 39 gestational weeks (A) and in one case of control subjects, a onemonth-old baby (B), positive for MBP stain , oligodendrocytes (thick arrows). x 200.
310 Brain & Development, Vol 13, No 5, 1991
DISCUSSION There are a number of reports on the myelination of the optic nerves [1, 3-19]. According to Duke-Elder and Cook [3], there was no evidence of the process which depended on oligodendrocytes until about the 20th gestational week when meylination was evident in the geniculate bodies and some time during the 24th gestational week it was recognizable in the tract. During the 24th-28th gestational week myelination reached the chiasma, in 32 weeks it was usually apparent in the optic nerve, and it slowly extended distally to reach the level of the lamina cribrosa at birth or some weeks thereafter. Also in the reports by Magoon and Robb [4] and Nakayama [5,6], the myelination began in the 28th-32nd gestational week. On the other hand, there were reports by Bernheimer [7] and Sattler [8] that they reported myelination in the optic tracts but not optic nerves of fetuses aged 28-32 gestational weeks under light microscope. As a result of our present investigation, like the former reports, MBP-positive cells were found in the optic nerves of the fetus at 32 weeks. This result indicates that the myelination began about the 32nd gestational week. It was supported by the fact that we examined 23 embryos and fetuses of 6-28 weeks and MBP was negative in all the cases examined. With regard to the time of appearance of glial cells, Nakaizumi [9] examined 60 cases of human fetuses aged up to 28 weeks histologically. He stated that the glial cells developed from the "inner layer" of the optic stalks after 12 gestational weeks. Raff et al [20] examined cell suspensions and cultures of optic nerves of rats and summarized that primitive glial cells develop as two distinct lineages which diverge to form astrocytes and oligodendrocytes prenatally. David et al [21] provided evidence that the rat optic nerve contains three types of macroglial cells that develop as two distinct lineages: one
(mm) 3
< Growth of Optic Nerve>
gestational weeks
Fig 5 There is a rapid increase in the size of optic nerve after 24 gestational weeks.
lineage comprises type 1 astrocytes, which develop before birth, while the other comprises oligodendrocytes and type 2 astrocytes, which develop after birth from a common, bipotential glial progenitor cell. GF AP-potitive cells were also observed in our present investigation in a case at 24 gestational weeks. This means that around this stage, as David et al [21] reported, one lineage comprises type 1 astrocytes and also, as Kuwabara [10] stated, the neuroepithelial cells of the fissure portion of the optic stalks are eliminated by marked bionecrosis. It is to be considered that it is at this time that a small number of the surviving cells differentiate into type 1 astrocytes, and that they divide the nerve fibers into small axon bundles and participate in the development of the optic nerves by surrounding them. In fact, during the period after 24 weeks till the time when MBP positive cells appear, the optic nerves increase their territory at a speed of twice as fast as before. Also after the appearance of MBP-positive cells, the rate of development increases twice as fast as before (Fig 5). Following the increase of the cross sectional area of the optic nerve, MBP-positive cells and GF AP-positive cells appear, and gradually the numbers of these cells and the positivity of stained nerve fibers increases. This fact indicates that these cells participate in the full growth of the optic nerves. Recently, Choi et al [22-24] suggested that radial glias may give rise to both astrocyte and oligodendrocyte with correlative use of immunohistochemical methods, electron microscopy and Golgi methods in the human fetal spinal cord at 6 to 17 weeks as well as in human fetal cerebrum and cerebellum. They showed that "transitional forms" between astrocyte and oligodendrocyte may exist. These "transitional forms" cells had mutually exclusive identifying features of astrocytes and oligodendrocytes respectively under electron microscope and demonstrated reverse immune reaction for MBP or GFAP on the contrary to their morphological features. In this report, further discussion to the peripheral or central developing theories of the optic nerve is out of the main problem since precise examination to the whole optic nerve is not done here. However, if radial glia or radial glial cell-derived astrocyte actually undergo differentiation or transformation into myelin-forming oligodendrocyte after the release of some still undetermined Signal in accordance with the "critical diameter" of axon as Choi et al reported, it can be suggested that the myelination does not necessarily occur from the center to the periphery. Their hypothesis may be supported by the fact that MBP-positive cells increased in number and crowded here and there, where nerve fibers became stained diffusely to be seen in the optic nerve of case No 26. As a result of our present study, MBP and GFAP immunostains were both negative until 21 gestational weeks, and the differentiation into astrocytes could not be ob-
Takayama et al: Immunohistochemistry of optic nerve 311
served. As reported by Silver and others [25] it was considered that this was the stage when axons invaded the intercellular spaces and created marked bionecrosis of many neuroepithelial cells in the fissure portion of the optic stalk. Probably before this stage, the morphological and functional growth of the nerve fibers was not achieved. Among the studied embryos younger than 24 weeks, some cases with strong tissue dissolution were included, and there is the possibility that the characteristics of antigen were lost due to this and left no stains. Moreover, there was a report by Cravioto [1] that he observed the optic nerves in 14-22 gestational weeks using an electron microscope, and that he confirmed the obvious myelination. Taking these factors into consideration, a careful selection of cases and an investigation in detail should be our future research. ACKNOWLEDGMENTS The authors wish to express their sincere gratitude to Dr. J. Takenouchi, Meimai Central Hospital, and Dr. K. Fukutome, Wakamiya Hospital for their kind cooperation for this research. REFERENCES 1. Cravioto H. Electron microscopic studies on the developing human nervous system. II. the optic nerves. J Neuropathol E'xpNeuroI1965;24:164-7. 2. Iwamoto K. Mean weeks of fetal life according to each calculation of CRL (in Japanese). Nihon Sanka Fujinka Gakkai Zasshi (Tokyo) 1983;35:2330. 3. Duke-Elder S, Cook C. System of ophthalmology. Vol III. London: Henry Kimpton, 1963: 109. 4. Magoon EH, Robb RM. Development of meylin in human optic nerve and tract: a light and electron microscopic study. Arch aphthalmol 1981;99:655-9. 5. Nakayama K. Studies of the myelination of the human optic nerve. Jpn J aphthalmol (Tokyo) 1968; 11 :132-40. 6. Nakayama K. Studies on the human optic nerve, especially its myelinization (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1967;70:1511-25. 7. Bernheimer S. Uber die Entwicklung und den Verlauf der Markfasern in Chiasm nervorum opticorum des Menschen. Arch Augenheilk 1889;20:133-79. 8. Sattler CH. Uber die Markscheidenentwicklung im Tractus opticus, Chiasm und Nervus opticus. Graefes Arch Qin Exp 1915;90:271-98.
312 Brain & Development, Vol 13, No 5, 1991
9. Nakaizurni Y. Embryological study of optic nerve of the Japanese embryo (in Japanese). Rinsho Ganka (Tokyo) 1958; 12:1329-42. 10. Kuwabara T. Development of the optic nerve of the rat. Invest aphthalmol Vis Sci 1975; 14:732-45. 11. Yamamoto T. Electron-microscopic observation of human optic nerves (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1965;69:1527-80. 12. Kanazawa S. Electron microscopic study of the human fetal optic nerve (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1970; 73 :1330-53. 13. Naumann GOH. Pathologie des Auges. Berlin: SpringerVerlag, 1980. 14. Friede RL, Hu KH. Proximo-distal differences in myelin development in human optic fibers. Z Zellforschung 1967; 79:259-64. 15. Kennedy PGE, Lisak RP, Raff MC. Cen type-specific markers for human glial and neuronal cells in culture. Lab Invest 1980;43:342-51. 16. Quitschke W, Sibony P, Schechter N. Variable expressio'n of intermediate filament proteins during embryonic development of human optic nerve. Exp Neurol 1985;90:204-14. 17. Okuda T. The development of myelin formation in the human optic nerve (in Japanese). Nihon Ganka Gakkai Zasshi (Tokyo) 1985;89:1156-65. 18. Shirai S. Embryology of optic nerve (in Japanese). Shinkei Ganka (Kanagawa) 1988;5:244-50. 19. Peters A. Observation on the connections between myelin sheaths and glial cells in the optic nerve of young rats. J Anat 1964;98: 125-34. 20. Raff MC, Abney ER, Miller RH. Two glial cell lineages diverge prenatally in rat optic nerve. Dev Bioi 1984;106: 53-60. 21. David S, Miller RH, Patel R, Raff MC. Effects of neonatal transection on glial cell development in the rat optic nerve: evidence that the oligodendrocyte-type 2 astrocyte cell lineage depends on axons for its survival. J Neurocytol 1984; 13:961-74. 22. Choi BH, Kim RC. Expression of glial fibrillary acidic protein in immature oligodendroglia. Science 1984;223:407-8. 23. Choi BH, Kim RC, Lapham LW. Do radial glia give rise to both astroglia and oligodendroglial cells? Dev Brain Res 1983;8:119-30. 24. Choi BH. Radial glia of developing human fetal spinal cord: Golgi, immunohistochemical and electron microscopic study. DevBrainRes 1981;1:249-67. 25. Silver J, Robb RM. Studies on the development of the eye cup and optic nerve in normal mice and in mutants with congenital optic nerve aplasia. Dev Bioi 1979 ;68: 175-90.
