quantification of myelin basic protein in the human fetal spinal cord during the midtrimester of...

THE JOURNAL OF COMPARATIVE NEUROLOGY 376:306-314 (1996)

Quantification of Myelin Basic Protein in the Human Fetal Spinal Cord During

the Midtrimester of Gestation

WILLIAM E. GREVER, FUNG-CHOW CHIU, MARIANELA TRICOCHE, WILLIAM K. RASHBAUM, KAREN M. WEIDENHEIM, AND WILLIAM D. LYMAN

Departments of Pathology, Neurology, and Obstetrics and Gynecology, Albert Einstein College of Medicine, Bronx, New York 10461

ABSTRACT The amount of myelin basic protein (MBP) was quantified in human fetal spinal cords from

12 to 24 gestational weeks (GW). MBP expression was determined by Northern blot, quantitative immunoblot, and immunocytochemistry. The development of compact myelin was analyzed by electron microscopy. Thirty-eight human fetal spinal cords were obtained after elective termination of intrauterine pregnancies from healthy women. Northern blot analysis showed a 15.8-fold increase in MBP mRNA between 12 and 18 GW. From 18 to 24 GW, MBP mRNA increased by 2.2-fold. The mRNA data paralleled immunoblot results that showed a 90.5-fold increase in MBP (0.147 ng/mg to 13.3 ng/mg tissue) between 12 and 18 GW and an approximately 11.5-fold increase between 18 and 24 GW (13.3 ngimg to 154 ng/mg tissue). Immunocytochemical analysis also showed increased staining for MBP with advancing gestational age. At 12 GW, MBP immunoreactivity was observed in all three spinal cord funiculi. By 18 GW, MBP was expressed throughout the spinal cord white matter with the exception of the lateral corticospinal tracts and in the rostra1 levels of the fasciculus gracilis. With respect to myelin, at 12 GW, rare, noncompacted myelin lamellae were observed by electron microscopy. By 18 GW, discrete areas of compact myelin were observed in areas that showed MBP immunoreactivity, and at 24 GW, compact myelin was prominent throughout the white matter of the spinal cord. This study demonstrates a quantitative increase in MBP expression that is associated with myelin formation during, the second trimester of human gestation. This information may provide normative data that can aid in the diagnosis of myelin disorders of the preterm, neonatal, and pediatric spinal cord. o 1996 Wiley-Liss, Inc.

Indexing terms: central nervous system, gene expression, human development, northern blotting, immunoblotting

Myelin basic protein (MBP) expression in the central nervous system (CNS) has been studied in many animal species (Campagnoni et al., 1987; Gordon et al., 1990). The formation of myelin in the spinal cord has also been documented for several of these species, including rat (Macklin et al., 1983-84; Schwab and Schnell, 1989), mouse (Choi, 1986), opossum (Ghooray and Martin, 1993), rabbit (Gillespie et al., 19901, and chicken (Macklin and Weill, 1985). The normal synthesis of myelin membrane and the compaction of myelin lamellae require the coordinated expression of several oligodendrocyte-specific genes. The sequential expression of these genes and the maturation of oligodendrocytes has been studied in dissociated cell cul- ture. For example, oligodendrocyte-enriched cultures from neonatal rodent optic nerve and brain have used to show a correlative relationship between the level of oligodendrocyte maturation and the expression of MBP, proteolipid

protein, myelin-associated glycoprotein, 2,3-cyclic nucleo- tide phosphodiesterase, carbonic anhydrase, protein kinase C isozymes, and various glycolipids (Cammeron and Rakic, 1991; Campagnoni et al., 1991; Skoff and Knapp, 1991; Asorta and Macklin, 1993, 1994; Gard and Pfeiffer, 1993).

Although studies have been published that compare neurogenesis and neurodevelopment between rodent and man (Bayer et al., 19931, the progression of oligodendrocyte development and the formation of myelin in the midtrimester human fetus remain to be established in detail (Sasaki et al., 1988; Martinez, 1989; Takayama et al., 1991; Ha-

Accepted August 21, 1996 W.E. Grever's current address is University of Wisconsin-Madison, School

of Veterinary Medicine, Department of Medical Sciences, 2015 Linden Drive West, Madison, WI 53706. E-mail: [email protected]. Address reprint requests there.

o 1996 WILEY-LISS, INC.

QUANTIFICATION OF MYELIN BASIC PROTEIN

segawa et al., 1992; Ozawa et al., 1994). This is especially true with respect to the onset of myelination in the human spinal cord. Little is known about the ontogeny of oligodendrocytes in the human spinal cord and the temporal and spatial expression of myelin-specific genes between 10 and 24 gestational weeks (GW) (Okado, 1980, 1982; Kronquist et al., 1987; Tohyama et al., 1991; Weidenheim et al., 1992, 1993; Bodhireddy et al., 1994).

For the present study, we measured the relative amounts of MBP messenger RNA (mRNA) and the absolute quantities of MBP in the developing human spinal cord. These values are compared with the temporal and spatial formation of myelin. The results are presented in conjunction with morphological data that describe the localization of MBP and the formation of compact myelin in the human fetal spinal cord between 12 and 24 GW.

307

isolated according to the manufacture’s instructions. Ap- proximately 11 pg of RNA was loaded per lane for the 12, 18, and 24 GW and adult human spinal cords. Standard techniques were used for the Northern gels (Sambrook et al., 1989). The RNA was transferred to a positively charged nylon membrane and UV cross-linked. The Northern blot was hybridized in Rapid-hyb buffer (Amersham, Arlington Heights, IL). The probe for MBP mRNA was washed at stringency of 0.1 x standard saline citrate (SSC). To control for the amount of RNA loaded for each sample, the Northern blot was stripped and re-probed for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA and washed at a stringency of 0.1 x SSC.

Densitometry for the Northern blots was performed using X-AR film (Kodak, Rochester, NY) and a Molecular Dynamics computing densitometer. To compensate for any variation in the amount of RNA between samples loaded onto the gel, ratios were calculated for the densitometric values of the MBP band and the GAPDH band for each sample.

Five micrograms of poly A+ RNA for three negative controls was analyzed on a separate Northern blot (data not shown). The negative controls consisted of a neuroblastoma cell line (MSN), cultured human fetal cortical astrocytes, and leptomeninges from human fetal brain. Human fetal astrocytes were cultured by using published techniques (Hurwitz et al., 1992).

Probes Probes were generated from agarose gel purified plasmid

inserts of cDNA for the respective proteins. [ C X - ~ ~ P I ~ C T P - labeled random primed probes were generated, and unincor- porated nucleotides were removed by size exclusion chromatography. The MBP probe was generated from a 1.2-Kb human cDNA sequence for the 21.5-kDa isoform (Roth et al., 1987) and added to a specific activity of 3.1 x lo6 cpm/mL. The GAPDH probe was generated from a 1.3-Kb human cDNA sequence (Ercolani et al., 1988) and added to a specific activity of 2.0 x lo6 cpm/mL.

Preparation of MBP standard For the purpose of quantifying the amount of MBP in

fetal spinal cord, a purified standard of adult MBP was generated. Normal human adult white matter was homogenized in ch1oroform:methanol (2:1, v:v). The extract was precipitated in 0.1 N HC1 and filtered. The filtrate was collected, concentrated by lyophilization, and dissolved in 8 M urea. The acid-extracted protein was purified by fast protein liquid chromatography (FPLC) on a Mono S HR 5/5 column (Pharmacia, Piscataway, NJ), and the fractions containing the major elution peak were pooled. The MBP standard was separated by sodium dodecylsulfate-polyacril- amide gel electrophoresis (SDS-PAGE) and stained with Coomassie Blue. The Coomassie Blue bands overlapped completely with the distribution of MBP immunoreactivity as demonstrated by immunoblotting. The MBP standard was quantified by the method of Bradford (Bradford, 19761, mixed with 0.1% bovine serum albumin, aliquoted, and stored at -80°C.

Immunoblot analysis of MBP To extract MBP from human fetal spinal cords, samples

of the same gestational age were weighed, pooled, thawed, and homogenized in ch1oroform:methanol. The extract was precipitated in 0.1 N HC1, concentrated, dissolved in 8 M

MATERIALS AND METHODS Tissue acquisition

These studies were approved by the Albert Einstein College of Medicine Committee on Clinical Investigation. Tissue was collected within 10 minutes of fetal demise after elective termination of intrauterine pregnancies from healthy women. Multiple parameters were used to determine gestational age. These included the date of the last menstrual period, uterine size as determined by bimanual and abdominal examination, ultrasonography using pre- dominantly the maximum biparietal diameter, and by physical measurement of fetal foot length after the termination of pregnancy (Hern, 1984).

For Northern and immunoblot analyses, human fetal spinal cords were transected at the superior aspect of the cervical enlargement and dissected from the vertebral column. After removal of dura matter, spinal cords, including all segments from the cervical enlargement to the conus medullaris, were placed in cryogenic vials and immediately frozen in liquid nitrogen. A segment of adult human thoracic spinal cord was obtained from an autopsy of a 67-year-old subject who died of a non-neurological disease. Whole mouse spinal cords were removed by insufflation, placed in cryogenic vials, and immediately frozen. Spinal cord tissue was stored at -80°C until used for mRNA or protein analysis.

For light and electron microscopic analyses, the cervical region was isolated immediately after dissection from the vertebral column and immersed in Bouin’s solution or Trump’s fixative as previously described (Weidenheim et al., 1992, 1993).

The gestational age and number of spinal cords analyzed by each procedure are shown in Table 1. A total of 38 human fetal spinal cords were available for this study.

Northern blot analysis of MBP mRNA Frozen spinal cords of the same gestational age were

combined and homogenized in TRI REAGENT (Molecular Research Center, Inc., Cincinnati, OH). Total RNA was

TABLE 1. Number of Spinal Cords per Procedure

Fetal age Northern Immuno- Ultrastructural (weeks) blot Immunoblot cytochemistry analysis

12 4 6 5 1 18 6 2 4 1 24 3 2 3 1

W.E. GREVER ET AL. 308

TABLE 2. Quantification of Myelin Basic Protein (MBP) in Human Fetal Spinal Cord

Total wet Average wet Fetal age weight of weight per Total acid- MBPIpg MBPImg

combined spinal cord extracted acid-extracted Total wet weight MBPispinal (gestational cord (ngI4 weeks) Number tissue (mg) (mg) protein (fig) protein (ng)l MBP (ng)* (nd3

18 2 839.7 419.9 1,085 10.31 11,181 13.3 5,590 24 2 2,420 1,210 7,000 53.07 371,513 153.5 185,757

12 6 399.8 0.66 112 0.53 58.8 0.147 9.8

'Determined by quantitative immunoblots f Fig. 2a). ZCalculated by multiplying the amount (bg) of acid-extracted protein by the quantity of MBP per g of acid-extracted protein. 3Calculated by dividing the total MBP (ng) obtained from a spinal cord preparation by the starting wet weight (mg) of the tissue. 4Calculated by multiplying the amount of MBP (ng) per milligram ofwet weight by the average wet weight (mg) of a fetal spinal cord for that gestational age.

urea, and the protein was quantified. For experiments using total protein, human fetal and adult spinal cords and adult mouse spinal cords were homogenized in 8 M urea.

Immunoblots were performed by using standard methods (Towbin et al., 1979). For quantitative immunoblot analysis, the amount of acid-extracted fetal protein was compared to a standard curve generated from 5 , 10,20,30, and 60 ng of the purified MBP standard. Lanes containing the purified MBP standard and all the fetal samples were contained on the same immunoblot (Chiu et al., 1988). The amount of acid-extracted protein loaded on the gel for each fetal sample is as follows: 12 GW 15,20,25 ng; 18 GW: 1.0, 2.0,3.0 ng; 24 GW: 0.25,0.50, 1.0 ng. Three concentrations of acid-extracted protein were analyzed to assure that the results would be in the linear range of the purified MBP standard curve. After SDS-PAGE, proteins were transferred to nitrocellulose and cross-linked to the membrane with N-Hydroxysuccinimidyl-4-azidobenzoate (HSAB; Pierce; Rockford, IL) by using the method of Kakita et al. (1982).

The nitrocellulose sheets were blocked in Tris-buffered saline (TBS) + 0.1% Tween-20 (TBST) containing 5% nonfat dry milk and incubated with a monoclonal anti-MBP antibody (Boehringer-Mannheim Biochemicals, Indianapo- lis, IN) diluted 1:200 in TBST + 5% milk. The MBP- immunoreactive bands were visualized using goat anti- mouse antibodies conjugated to horseradish peroxidase (Bio-Rad Laboratories, Richmond, CA) and enhanced chemiluminescence (ECL; Amersham, Arlington Heights, IL). The combined densitometric value of all immunoreactive bands in each lane was determined by analysis of X-AR film on a Molecular Dynamics computing densitometer. The immunoreactive-densitometric value from each of the three acid-extracted protein concentrations was compared to the amount of protein loaded. When the densitometric value was plotted against the amount of protein loaded on the gel, the result was a straight line for each of the three fetal samples. The reliability of the quantitative immunoblot technique was evident in the linear correlation coefficient (9) for the densitometric value vs. loaded protein plots. The r2 values for the 12, 18, and 24 GW samples were 0.946, 0.999, and 0.997, respectively, with 1.0 being a perfect linear correlation. The average of the three densitometric values for each fetal sample was used to calculate the

fetal tissue, spinal cord tissue was homogenized in 8 M urea. The adult human spinal cord tissue was a 2-cm segment of thoracic spinal cord frozen after a 72-hour postmortem interval. Whole spinal cords were used for the mouse and 18 GW human samples. These 18 GW samples are not included in Tables 1 or 2.

Immunocytochemistry Bouin's-fixed material was equilibrated in 30% sucrose,

and 40-*m Vibratome sections were cut as previously described (Weidenheim et al., 1992, 1993). The same anti- MBP antibody used for immunoblot analysis was employed for the immunocytochemical studies used at dilutions of 1:500-1:1,600. MBP immunoreactivity was detected with secondary antibodies coupled to horseradish peroxidase and

MBP standard was calculated by using the AssayZap program (ver. 1.41) for the Macintosh computer (Biosoft, Cambridge, UK). This program was also used to calculate the amount of MBP in each fetal sample as determined against the standard curve.

To obtain total Protein from spinal cord tissue used in the comparison of mouse, adult human, and 18 GW human

ofhuman myelin-basic protein (MBP). The positive band was approximately 1.5 Kb. B: Hybridization to probe generated from cDNA for human glyceraldehyde-3-phosphate dehydrogenase (GAPDH). This probe hybridized to a band at approximately 0.9 Kb. All lanes contained approximately 11 pg of total RNA from human spinal cords. The normalized MBP to GAPDH mRNA ratios for the 12 gestational week (GW), 18 GW, 24 GW, and adult human spinal cords are 0.03,0.47, and 1.0, and 1.69, respectively.


a Purified MBP Standard

309

12 GW 18 GW 24 GW

29-

5 10 20 30 60 15 20 25 1.0 2.0 3.0 0.25 0.50 1.00 nanograms micrograms

b

31 - 29-

20-

18-

Fig. 2. Quantitative immunoblots of MBP from human fetal spinal cords. A Enhanced chemiluminescence (ECL) film from the immunoblot of purified MBP standard and protein from human fetal spinal cords. The immunoreactive doublet in the MBP standard represents the 18.5- and 17-kDa isoforms of MBP found in mature white matter. The quantity of MBP detected in the fetal spinal cord samples are shown in Table 2. B: ECL film from MBP immunoblot of total protein from adult mouse, adult human, and 18 GW human fetal spinal cords. Immunodetection of MBP isoforms can be compared between samples. A comparison can be made between the MBP isoforms isolated in the 18 GW total protein preparation and the 18 GW acid-extracted preparations in lane 10 of A. Tshe position of molecular weight markers are shown in kDa.

visualized with 3,3' diaminobenzidine (DAB; Weidenheim et al., 1992, 1993).

in Epon, cut at 1 pm thickness, and stained with toluidine- blue. Thin sections from the dorsal column of the cervical

Ultrastructural analysis segments were processed for ultrastructural study by staining with lead citrate and uranvl acetate. Examination of

The lumbosacral regions of fetal spinal cords initially fixed in Trump's solution were postfixed in Os04 embedded

dorsal column thin sections was performed on a Siemens 102 electron microscope.

W.E. GREVER ET AL.

Gestational Weeks

Fig. 3. MBP gene expression and fetal spinal cord mass vs. gestational weeks. The average mass of a fetal spinal cord is shown by the dashed line plotted against the left vertical axis. Circles indicate the ratios of MBP mRNA to GAPDH mRNA. The mRNA ratio data were normalized with respect to the value obtained at 24 gestational weeks, and are plotted against the scale on the left vertical axis. The squares show the quantity in micrograms of MBP per spinal cord. Triangles represent the quantity of MBP in nanograms per milligram of tissue. Note that the MBP per spinal cord and the MBP per milligram tissue are both plotted against the right vertical axis, and the scale is to be read in micrograms and nanograms, respectively.

Electronic imaging of micrographs The micrographs of anti-MBP-labeled spinal cord sec-

tions in Figure 4 were generated from negatives of black and white print film scanned in a 35-mm slide scanner. Figure 5 was generated by scanning 8 x 10 inch prints of electron micrographs on a flatbed scanner. Image contrast was adjusted using the computer program Photoshop (Adobe Systems Inc; Mountain View, CA) to provide a uniform appearance among panels in each figure.

RESULTS Measurement of MBP/GAPDH mRNA ratios The amount of MBP-specific message increased 15.8-fold

between 12 and 18 GW and increased by a factor of 2.2 between 18 and 24 GW (Fig. la). The ratios of MBP to GAPDH mRNA densitometric values for 12, 18, and 24 GW were 0.03, 0.47, and 1.0, respectively. The normalized MBP/GAPDH mRNA value for the adult spinal cord was 1.7. The intensity of the hybridization signal to GAPDH mRNA was equivalent for the fetal spinal cord samples, indicating consistent loading of RNA in these lanes (Fig. lb). To compensate for the weak MBP hybridization signal from the 11 pg of 12 GW RNA, another lane on the same northern blot had 46 pg of 12 GW RNA loaded. The MBP mRNA hybridization signal in this lane was increased compared to the lane with 11 pg RNA. The MBP-to- GAPDH ratios in the two 12 GW lanes were similar (data not shown). On a separate Northern blot, negative controls of a neuroblastoma cell line (MSN), cultured human fetal cortical astrocytes, and leptomeninges from human fetal brain were all negative for MBP mRNA.

Quantification of MBP MBP was detected in the 12 GW spinal cord samples, the

earliest gestational age examined. The amount of MBP in

acid-extracted protein from the 12 GW sample was 0.147 ng of MBP per milligram of tissue, or about 9.8 ng per spinal cord, as calculated from the standard curve of purified MBP on the same immunoblot (Fig. 2a). In the 18 GW sample, the intensity of the MBP bands corresponded to 13.3 ng of MBP per milligram of tissue, or 5.6 pg per spinal cord. At the latest gestational age sampled, 24 GW, there were 154 ng of MBP per milligram of tissue. This was equivalent to 186 pg of MBP per spinal cord (Table 2). This represents a 1,040-fold increase in the amount of MBP per unit mass of spinal cord, or a 19,000-fold increase in MBP per spinal cord as compared to the 12-week samples.

The isoforms of MBP detected in the fetal spinal cords were 21.5, 20.5, 18.5, and 17.2 kDa. However, a faint immunoreactive band was also detected at 14 kDa in the human fetal spinal cord samples. In a separate experiment, a gel and immunoblot with several sets of molecular weight markers were analyzed to estimate the apparent molecular weights of MBP isoforms in human fetal and adult spinal cords (Fig. 2b). Protein from adult mouse spinal cord was included on this immunoblot to compare the MBP isoforms in rodent species to the isoforms in human. This immunoblot can also be used to compare the isoforms of MBP between acid-extracted and total protein preparations from human fetal spinal cords. The acid-extraction procedure appeared to isolate all isoforms equally well (Fig. 2b).

MBP expression relative to spinal cord mass The average mass of a human fetal spinal cord increases

approximately 18-fold between 12 and 24 GW, from 67 mg to 1.21 g. The quantity of MBP per spinal cord detected by immunoblot analysis increased approximately 19,000-fold. When the amount of MBP per unit mass was compared between 12 and 24 GW there was a 1,044-fold increase. The increase in the ratio of MBPiGAPDH mRNA during this period of gestation was 34-fold. During the initial weeks of the midtrimester, the rates of increase in mass of the spinal cord and MBP/GAPDH mRNA ratio were greater than the rate of increase in the production of protein as determined by MBP immunoblot analysis. However, between 18 and 24 GW the rate of increase in detectable protein becomes greater than the rate of increase of spinal cord mass and the MBPiGAPDH mRNA ratio (Fig. 3).

Immunocytochemical localization of MBP At 12 GW, sparsely distributed “lacy” oligodendrocytes

and short myelin tubules could be identified in fasciculus gracilis, and the anterior and lateral funiculi of the cervical spinal cord by labeling with anti-MBP antibodies (Fig. 4a). Previous studies have shown a rostral-to-caudal gradient of immunoreactivity with more positive structures seen in superior levels of the spinal cord at this age (Weidenheim et al., 1992,1993). By 18 GW, MBP was expressed throughout the white matter in all three funiculi (Fig. 4b). Near the end of the second trimester, 24 GW, MBP was expressed in the same regions as 18 GW, but the staining was both more prevalent and more intense (Fig. 4c). The corticospinal tracts remained unlabeled at all ages examined in this study. The sparse MBP immunoreactivity present in the dorsolateral funiculus at 24 GW (Fig. 4c) is likely due to fibers not contained in the corticospinal tract.

Ultrastructural analysis of myelin lamellae Toluidine-blue stained 1-pm-thick sections of lumbosa-

cral spinal cord were examined for the presence of process-

QUANTIFICATION OF MYELIN BASIC PROTEIN 311

Fig. 4. Immunocytochemical localization of MBP in the cervical region of human fetal spinal cords. Forty-micron sections of fixed spinal cord were labeled with anti-MBP antibodies. MBP-positive structures were revealed by the enzymatic conversion of 3,3'-diaminobenzidine

(DAB). a-c show spinal cords of 12, 18, and 23-24 gestational weeks, respectively. As gestation progresses, there is an increase in MBP staining in the fasciculus gracilis, the anterior and lateral funiculi, and in the regions of gray matter. Scale bar = 0.5 mm.

bearing cells forming myelin tubules. At 12 GW, only a few cells were present in the areas of developing white matter. Ultrastructural examination of thin sections from this gestational age group revealed occasional glial processes enveloping axons, but compact myelin was not observed (Figs. 5a,b). By 18 GW, 1-km sections showed numerous cells and myelin tubules in the white matter, and ultrastructural examination confirmed the presence of thin lamellae of compact myelin (Figs. 5c,d). At 24 GW, myelin tubules were easily identified in 1-km sections. Ultrastructural analysis showed that myelin lamellae were more numerous and thicker than observed in the samples from earlier gestational ages (Fig. 5e,f).

DISCUSSION The results of this study show that there is a steady

increase in the expression of MBP measured by multiple parameters in human fetal spinal cords during the second trimester of gestation. This increase in MBP expression was not due to the developmental increase in spinal cord mass. Rather, it specifically reflects the differentiation of oligodendrocytes and development of compact myelin. This indicates that the white matter comprises a greater proportion of the total weight as the tissue develops. These conclusions are supported by the data from the Northern analysis of MBP mRNA, MBP immunoblot analysis, and the number of MBP-positive cells and tubules detected by immunocytochemistry throughout the second trimester. These measures were also in accordance with the progressive appearance of compact myelin lamellae in the developing white matter of the human fetal spinal cord.

I t is interesting to note that these correlations exist even though the Northern and immunoblot results are based on samples of whole spinal cord whereas the light and electron

microscopy are from the cervical region. This is to be expected. Myelination is initiated earlier in the cervical and thoracic than in the lumbar region of the human spinal cord. However, once initiated, myelination appears to progress at a similar rate for the three regions (Gilles et al., 1983). Therefore, visual examination of myelin in the cervical region may provide a high estimation of myelination within the spinal cord as a whole, but comparison of this region at 12, 18, and 24 gestational weeks, as done in this study, should provide an accurate estimate of the progression of myelination.

We have shown that the relative amount of MBP mRNA at 12 GW (the MBP/GAPDH mRNA ratio) is less than 3% of the MBP/GAPDH mRNA ratio observed at 24 GW and less then 2% of that ratio in the adult human. The quantitative results from the Northern analysis reported here agree favorably with the qualitative data of Kronquist et al. (Kronquist et al., 1987).

To document further the progression of MBP expression, quantitative immunoblot analysis was used to determine the amount of this protein in acid-extracted material. The acid-extraction procedure was applied to the fetal spinal cord samples to maintain consistency with the method of preparation used to purify the MBP standard. The results of the immunoblot analysis were that the amount of MBP in a 12 GW human fetal spinal cord is approximately 0.005% of that in a 24 GW spinal cord.

In addition to the amount of MBP detected per spinal cord, the immunoblot data suggest that there is a variation in the MBP isoforms expressed at different gestational ages. There appears to be an increase in the relative proportion of the 21.5-kDa isoform compared to the other isoforms as gestational age increases (Fig. 2b). There also appears to be a decrease in the relative amount of immunoreactivity detected around 14 kDa as gestation proceeds. I t is un-

312 W.E. GREVER ET AL.

Fig. 5. Ultrastructure of myelin lamellae in the cervical region of human fetal spinal cords. a-f are micrographs from spinal cords of 12, 18, and 24 gestational weeks, respectively. b, d, and f show details of the myelin lamellae in a, c, and d, respectively. The number of lamellae in

the immature myelin sheaths increases with gestational age. The inner loop of the myelinating oligodendrocyte process is evident in the micrographs from 18 and 24 GW spinal cords (d, 0. Scale bars = 2 bm in a, c, and e; 0.2 bm in c, d, and f.

known if the 16kDa immunoreactivity represents a MBP isoform known to exist in rodents (de Ferra et al., 1985; Takahashi et al., 1985) or is a degradative product from a

larger isoform. Interpretation of these data are based on the assumptions that MBP is extracted with an equal efficiency at all the fetal ages studied and that acid-extraction does


not preferentially exclude a specific isoform. These assumptions are difficult to confirm because a change in the relative proportion of an isoform could be representative of a developmental change in expression.

When all isoforms are combined, the increase in MBP detected by immunoblot appears to increase geometrically. However, the ratio of MBP to GAPDH mRNA appears to increase at a linear rate (Fig. 3). The reason for the disparity between the rates of increase for mRNA and protein are unknown, but the data are suggestive of post-transcriptional regulation of MBP expression in the developing human CNS.

Ultrastructural examination of fetal spinal cords gave further support to the results of the Northern and immunoblotting studies. Starting at 12 GW and continuing to 24 GW, EM observations indicated a progressive increase in the amount of compact myelin with gestational age. This also correlates positively with immunocytochemical data that showed an increase in the number of cells, processes, and myelin lamellae labeling positively for MBP.

Expression of MBP and its use as a measure of myelination are well established. Although MBP is thought to be synthesized in oligodendrocyte processes and, to a lesser extent in the cell body (Trapp et al., 1987; Amur-Umarjee et al., 1990; Ainger et al., 1993), it is transported to the inner loop of compact myelin, and its presence is associated with the development of compacted mature myelin. Despite these observations, there is controversy about which oligodendrocyte protein is the best surrogate marker for myelin formation. The probable answer to this dispute is that multiple markers should be used in order to obtain the most accurate measure of myelination. However, as we have demonstrated, the expression of MBP is a reliable marker for myelination. These conclusions are supported by data acquired through four independent means: Northern blot analysis, immunoblot analysis, immunocytochemistry, and ultrastructural analysis. The results and techniques presented in this communication may aid in the accurate analysis of subtle changes in myelin formation during normal and pathologic development of the human CNS.

313

ACKNOWLEDGMENTS We are grateful for the technical assistance with the

electron microscopy offered by Gloria Stephney and Yvonne Kress of the Ultrastructural Pathology Laboratory at the Albert Einstein College of Medicine. These studies were supported by United States Public Health Service grants MH 46815 and MH 47667 and National Institutes of Health Training Grant NIH5T32CA09060. Data in this paper are from a thesis to be submitted in partial fulfillment of the requirements for the Degree of Doctor of Philosophy in the Sue Golding Graduate Division of Medical Sciences, Albert Einstein College of Medicine, Yeshiva University.

LITERATURE CITED Ainger, K., D. Avossa, F. Morgan, S.J. Hill, C. Barry, E. Barbarese, and J.H.

Carson (1993) Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J. Cell Biol. 123:431- 441.

Amur-Umarjee, S.G., L. Hall, and A.T. Campagnoni (1990) Spatial distrihu- tion of mRNAs for myelin proteins in primary cultures of mouse brain. Dev. Neurosci. 12:263-272.

Asotra, K., and W.B. Macklin (1993) Protein kinase C activity modulates myelin gene expression in enriched oligodendrocytes. J. Neurosci. Res. 34:571-588.

Asotra, K., and W.B. Macklin (1994) Developmental expression of protein kinase C isozymes in oligodendrocytes and their differential modulation hy 4p-phorhol-12,13-dibuterate. J. Neurosci. Res. 39:273-289.

Bayer, S.A., J. Altman, R.J. Russo, and X. Zhang (1993) Timetables of neurogenesis in the human brain based on experimentally determined patterns in the rat. Neurotoxicology 14:83-144.

Bodhireddy, S.R., W.D. Lyman, W.K. Rashbaum, and K.M. Weidenheim (1994) Immunohistochemical detection of myelin basic protein is a sensitive marker of myelination in second trimester human fetal spinal cord. J. Neuropathol. Exp. Neurol. 53.144-149,

Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:24%254.

Cammeron, R.S., and P. Rakic (1991) Glial cell lineage in the cerebral cortex: a review and synthesis. Glia 4:124-137.

Campagnoni, A.T., B. Sorg, H.J. Roth, K. Kronquist, S.L. Newman, K. Kitamura, C. Campagnoni, and B. Crandall(1987) Expression of myelin protein genes in the developing brain. J. Physiol. (Lond.) 82:229-238.

Campagnoni, A.T., J.M. Verdi, A.N. Verity, S. Amur-Umarjee, and S. Byravan (1991) Posttranscriptional regulation of myelin protein gene expression. Ann. NYAcad. Sci. 633:178-188.

Chiu, F.-C., R.S. Sacchi, L. Claudio, S. Kohayashi, and K. Suzuki (1988) Coexpression of glial fihrillary acidic protein and vimentin in the central and peripheral nervous system of the Twitcher mutant. Glia 1:105-112.

Choi, B.H. (1986) Myelin-forming oligodendrocytes of developing mouse spinal cord: immunocytochemical and ultrastructural studies. J. Neuro- pathol. Exp. Neurol. 45:513-24.

de Ferra, F., H. Engh, L. Hudson, J. Kamholz, C. Pukett, S. Molineaux, and R.A. Lazzarini (1985) Alternative splicing accounts for the four forms of myelin basic protein. Cell 46721-727.

Ercolani, L., M.D. Florence, and M. Alexander (1988) Isolation and complete sequence of a functional human glyceraldehyde-3-phosphate dehydroge- nasegene. J. Biol. Chem. 263:15335-15341.

Gard, A.L., and S.E. Pfeiffer (1993) Glial cell mitogens hFGF and PDGF differentially regulate development of 04+GalC- oligodendrocyte pro- genitors. Dev. Biol. 159:618-630.

Ghooray, G.T., and G.F. Martin (1993) The development of myelin in the spinal cord of the North American opossum and its possible role in the loss of ruhrospinal plasticity. A study using myelin basic protein and galactocerebroside immunohistochemistry. Brain Res. Dev. Brain Res. 72:67-74.

Gilles, F.H., and E.C. Dooling (1983) Myelinated tracts: growth patterns. In The Developing Human Brain. Boston: John Wright-PSG Inc., pp. 117-183.

Gillespie, C.S., B.D. Trapp, D.R. Colman, and P.J. Brophy (1990) Distribu- tion of myelin basic protein and P2 mRNAs in rabbit spinal cord oligodendrocytes. J. Neurochem. 54: 1556-15561.

Gordon, M.N., S. Kumar, A. Espinose de 10s Monteros, S. Scully, M.S. Zhang, J. Huher, R.A. Cole, and J. de Vellis (1990) Developmental regulation of myelin-associated genes in the normal and myelin deficient mutant rat. Adv. Exp. Med. Biol. 265t11-22.

Hasegawa, M., S. Houdou, T. Mito, S. Takashima, K. Asanuma, and T. Ohno (1992) Development of myelination in the human fetal and infant cerebrum: a myelin basic protein immunohistochemical study. Brain Dev. 14:l-6.

Hern, W.M. (1984) Correlation of fetal age and measurements between 10 and 26 weeks of gestation. Obstet. Gynocol. 6326-32.

Hurwitz, A.A., W.D. Lyman, M.P. Guida, T.M. Calderon, and J.W. Berman (1992) Tumor necrosis factor alpha induces adhesion molecule expression on human fetal astrocytes. J. Exp. Med. 176:1631-1636.

Kakita, K., K. O’Connell, and M.A. Permutt (1982) Immunodetection of insulin after transfer from gels to nitrocellulose filters. Diabetes 311648- 562.

Kronquist, K.E., B.F. Crandall, W.B. Macklin, and A.T. Campagnoni (1987) Expression of myelin proteins in the developing human spinal cord: cloning and sequencing of human proteolipid protein cDNA. J. Neurosci. Res. 18:395-401.

Macklin, W.B., E. Oherfield, and M.B. Lees (1983-84) Electrohlot analysis of rat myelin proteolipid protein and basic protein during development. Dev. Neurosci. 6:161-168.

Macklin, W.B., and C.L. Weill (1985) Appearance of myelin proteins during development in the chick central nervous system. Dev. Neurosci. 7:170- 178.

Martinez, M. (1989) Biochemical changes during early myelination of the human brain. In P. Evrard and A. Minkowski (eds): Developmental Neurohiology. New York: Raven Press, Ltd., pp. 185-200.

314 W.E. GREVER ET AL.

Okado, N. (1980) Development of the human cervical spinal cord with reference to synapse formation in the motor nucleus. J. Comp. Neurol. 191 :495-5 13.

Okado, N. (1982) Early myelin formation and glial cell development in the human spinal cord. Anat. Rec. 202483-490.

Ozawa, H., A. Nishida, T. Mito, and S. Takashima (1994) Development of ferritin-positive cells in cerebrum of human brain. Pediatr. Neurol. IOt44-48.

Roth, H.J., K.E. Kronquist, N. Kerlero de Rosbo, B.F. Crandall, and A.T. Campagnoni (1987) Evidence for the expression of four myelin basic protein variants in the developing human spinal cord through cDNA cloning. J. Neurosci. Res. 17t321-328.

Sambrook, J., E.F. Fritsch, and T. Maniatis (1989) Molecular Cloning. A laboratory manual. Cold Spring Harbor, Ny: Cold Spring Harbor Press.

Sasaki, A., J. Hirato, Y. Nakazato, and Y. Ishida (1988) Immunohistochemi- cal study of the early human fetal brain. Acta Neuropathol. (Berl.) 76128-134.

Schwab, M.E., and L. Schnell (1989) Region-specific appearance of myelin constituents in the developing rat spinal cord. J. Neurocytol. 18t161- 169.

Skoff, R.P., and P.E. Knapp (1991) Lineage and differentiation of oligodendrocytes in the brain. Ann. N.Y. Acad. Sci. 633:48-55.

Takahashi, N., A. Roach, D.B. Teplow, S.B. Prusiner, and L. Hood (1985) Cloning and characterization of the myelin basic protein gene from

mouse: one gene can encode both 14 and 18.5 kD MBPs by alternate use ofexons. Cell 42139-148.

Takayama, S., M. Yamamoto, K. Hashimoto, and H. Itoh (1991) Immunohis- tochemical study on the developing optic nerves in human embryos and fetuses. Brain Dev. 13,307-312.

Tohyama, T., V.M. Lee, L.B. Rorke, and J.Q. Trojanowski (1991) Molecular milestones that signal axonal maturation and the commitment of human spinal cord precursor cells to the neuronal or glial phenotype in development. J. Comp. Neurol. 310285-299.

Towbin, H., T. Staehelin, and J. Gordon (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some application. Proc. Natl. Acad. Sci. U.S.A. 76:43504354.

Trapp, B.D., T. Moench, M. Pulley, E. Barbosa, G. Tennekoon, and J. Griffin (1987) Spatial segregation of mRNA encoding myelin-specific proteins. Proc. Natl. Acad. Sci. U.S.A. 84:7773-7777.

Weidenheim, K.M., Y. Kress, I. Epshteyn, W.K., Rashbaum, and W.D. Lyman (1992) Early myelination in the human fetal lumbosacral spinal cord: characterization by light and electron microscopy. J. Neuropathol, Exp. Neurol. 51t142-149.

Weidenheim, K.M., I. Epshteyn, W.K. Rashbaum, and W.D. Lyman (1993) Neuroanatomical localization of myelin basic protein in the late first and early second trimester human feotal spinal cord and brainstem. J. Neurocytol. 22507-516.

quantification of myelin basic protein in the human fetal spinal cord during the midtrimester of...

Documents