Quantitative evaluation of axonal regeneration by immunochemical assay for neurofilament protein

EXPERIMENTAL

NEUROLOGY

100,83-97

(I 988)

Quantitative Evaluation of Axonal Regeneration by lmmunochemical Assay for Neurofilament Protein LLOYDGUTH,R.WAYNEALBERS,CHARLESP. AND EDWARD

BARRETT,

J. DONATI’

Department ofAnatomy, University ofMaryland School of Medicine, Baltimore, Maryland 21201; and Laboratory ofNeurochemistry. National Institute ofNeurological and Communicative Disorders and Stroke, National Institutes ofHealth, Bethesda, Maryland 20205 Received August 17.1987 In experiments on nerve regeneration requiring assessment ofthe rate and extent of axonal outgrowth, the availability of a simple and accurate method of quantification would be extremely useful. We approached this issue by modifying the conventional ELISA procedure so as to provide a sensitive, specific, and quantitative biochemical assay of the phosphorylated neurofilament content of homogenates or sections of nerve tissue. The technique involves four sequential steps: (i) adhesion of fixed or fresh homogenates or tissue sections to wells of microtiter plates, (ii) binding of a monoclonal antibody against phosphorylated neurofilament to the tissue, (iii) secondary binding to the anti-phosphorylated neurofilament of a phosphatase-labeled second antibody (antimouse IgG), and (iv) enzymatic assayof alkaline phosphatase activity using a fluorescent substrate (4-methylumbelliferyl phosphate). The technique is sufficiently sensitive to measure the phosphorylated neurofilament content of a 1:100,000 (w/v) homogenate of brain, spinal cord, or peripheral nerve and of single IO-pm paraffin sections of Bouin-fixed rat spinal cord. To validate the applicability of the procedure to the study of nerve regeneration, the sciatic nerve of adult rats

Abbreviations: CNS, PNS-central, peripheral nervous system; PNF-phosphorylated neurofilament. ’ The authors are indebted to Dr. Ludwig A. Stemberger and Dr. Nancy H. Stemberger for their helpful advice and for generously providing the antibodies used in these experiments. We are pleased to acknowledge the participation of Mr. Nicolas Andrews and Mr. Michael Rauser, who were supported by fellowships from the Dean’s Medical Student Research Fund of the University of Maryland School of Medicine. We also thank Mr. Fred Bland and Mr. Miki Kasai for their expert technical help. The research was funded by a grant from the National Institutes of Health (NS-2 1460). 83 0014-4886/88 $3.00 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.

84

GUTH

ET AL.

was either crushed (to permit regeneration) or transected and ligated (to preclude regeneration). The animals were autopsied 1 to 16 weeks later, when four segments 3-mm in length taken from regions proximal and distal to the lesion site were assayed for phosphorylated filament content. The temporal course of its disappearance during degeneration and its reappearance during regeneration coincided with the known histologic changes in crushed and transected nerves. These findings demonstrate the validity of using the immunochemical assayfor PNF in studies of nerve regeneration in the peripheral nervous system and the potential applicability of this procedure to studies on regeneration in the central nervous system. @ 1988 Academic hs.3, Inc.

INTRODUCTION Although for many years it was generally thought that regeneration in the mammalian central nervous system was extremely limited, the results of numerous recent experiments involving grafting of peripheral (1) and central (2) nervous tissues into the central nervous system (CNS) indicate that the grafted tissues are extensively reinnervated by host axons. Consequently, neuroscientists are increasingly inclined to view that regeneration in the mammalian CNS can occur provided that extrinsic neuronotrophic factors are available for stimulation of cell growth (30) and that an appropriate microenvironment is available to both guide the growing axons and support their continued elongation ( 1,9, 19). Current research approaches to the study of neuronal responses to trauma include (i) fundamental studies designed to elucidate the intrinsic cellular and extrinsic neurotrophic molecular factors that regulate the process of axonal elongation (17, 24, 29, 30) and (ii) applied studies in which therapeutic procedures are evaluated histologically for their effect on injured brain or spinal cord (5,7, 10, 11). In both cases, quantitative evaluation of the extent or rate of regeneration is often critical to the success of the experiments. Such quantitative evaluations are usually done by computerized image analysis of histologic preparations (15, 16, 32), but the accuracy of image analysis is restricted by the quality of the histologic preparations and the contrast provided by the staining technique. Although these deficiencies can be minimized by electron microscopical image analysis, this approach is not well suited to the survey of large pieces of tissue or large numbers of specimens. Consequently, the availability of a rapid and highly specific microchemical assay for nerve fibers would facilitate both basic studies on nerve regeneration as well as applied studies on the treatment of central and peripheral nerve inj?uies. We chose to adapt the enzyme-linked immunoabsorbent assay (ELISA) for this purpose, because this technique provides a simple, sensitive, and selective method for assaying minute concentrations of specific proteins in tissues. Furthermore, the ELISA method has been successfully applied to the quantitative analysis of neurofilament proteins in cultured nerve tissue (6).

AXONAL

REGENERATION

85

Neurofilaments are a major class of intermediate filament making up the cytoskeleton of neurons (22). They are composed of a protein core domain and triplet polypeptide subunits of 70, 150, or 200 kDa, the latter two being highly phosphorylated (3,4, 14). Furthermore, immunocytochemical studies with monoclonal antibodies that distinguish between the phosphorylated and nonphosphorylated forms of neurofilaments have shown that these polypeptides are specifically distributed within the central and peripheral nervous systems (27). In addition, neurofilaments disappear from degenerating axons, reappear during regeneration, and their concentration in the axon correlates with the axonal diameter (13). Because monoclonal antibody against the phosphorylated 200-kDa subunit (PNF) reacts immunocytochemically with peripheral and central axons rather than neuronal perikarya (20), this antibody should be a suitable marker for the quantitative, as well as qualitative (immunocytochemical), analysis of regeneration in the peripheral and central nervous systems. In the present contribution, we describe an ELISA procedure sufficiently sensitive to measure PNF quantitatively in a single IO-pm paraffin section or in 50 ~1 of a 1: 100,000 homogenate (w/v) of rat spinal cord. In addition, we validate the usefulness of this method for the study of axonal regeneration by comparing changes in PNF in the transected or crushed sciatic nerve of the rat. We show that the content of PNF decreases markedly within the first 2 weeks after injury and that it gradually increases in those animals in which regeneration is permitted but not in those in which regeneration is precluded. This quantitative procedure is currently being applied to the evaluation of regeneration in the injured spinal cord. MATERIALS

AND

METHODS

ELISA Assay Procedure. (a) Sample preparation: Nerve tissues were weighed and homogenized by hand in a ground glass homogenizer at 3°C in an appropriate volume of 0.0 1 mM sodium azide containing 1.O mM EDTA (to inhibit calcium-activated proteolysis). The resulting homogenate was diluted to 1: 10,000 (w/v) and serially diluted in the same medium to yield a range of dilutions for assay between 1: 10,000 and 1:200,000 (w/v). (b) Plating: From the diluted homogenates, 50-~1 samples were placed in the wells of polystyrene, 96-well plates (Nunc Inter Med, Denmark). (c) Drying: The samples were dried 60 min by placing the tray 15 in. beneath a hair dryer to provide a temperature of 35°C at the surface of the tray. (d) Blocking: Wells were filled with 250 ~1 of a 1:200 dilution of normal goat serum in K&P Diluent Solution (Kierkegaard and Perry, Gaithersburg, Maryland) for 60 min at 37°C. (e) Rinsing: Wells were rinsed gently twice with deionized water. (f) First antibody: A monoclonal antibody to PNF was diluted 1:lOOO in

86

GUTH

ET AL.

K&P Diluent Solution; 80 ~1 were pipetted to each well and incubated 60 rnin at 37°C. Wells were then washed thoroughly three times in a high ionic strength buffered salt solution (0.5 A4 NaCl, 0.05% Tween-20,0.0 1 M TrisHCl, pH 7.5) and 12 times in running deionized water. (g) Second antibody: K&P phosphatase-conjugated goat antimouse-IgG was diluted 1:200 in K&P Diluent Solution. Eighty microliters was added to each well and incubated 60 min at 37°C. Wells were washed thoroughly as in the previous step (three times in high salt solution and 12 times in running deionized water). (h) Enzyme assay: The tray was chilled to 3°C and 100 ~1 buffer-substrate was added to each well (0.2 mM methylumbelliferyl phosphate, 1 rnJ4 magnesium chloride, 100 udiethanolamine-HCl buffer, pH 8.8). The tray was incubated at 25 or 37°C and fluorometric readings were made at Time 0 and at IO-min intervals thereafter for 60 min (MicroFLUOR Reader, Dynatech Laboratories, Chantilly, Virginia). Protein Determination. Protein measurements were made on the undiluted homogenates using the microtiter plate adaptation ( 18) of the bicinochinic acid (25) procedure. Reagents were obtained from Pierce Chemical Co. (Rockford, Illinois). Optical density measurements were obtained at 570 nm using a MicroELISA Minireader (Dynatech Laboratories, Inc., Chantilly, Virginia) and compared with bovine serum albumin standards. The albumin solutions were always freshly prepared and treated simultaneously with each experimental assay. Preparation of Neurofilaments. The following procedure is an adaptation of Schlaepfer’s method (2 1,23). Two rats weighing 650 g each were anesthetized with chloral hydrate. The brains and spinal cords were removed and combined for a total weight of 6.5 g. The tissue was minced with scissors into 1-mm fragments and placed in six 15-min changes of a hypotonic medium (2.5 mMEDTA, 0.01 mMsodium azide, 0.2 mg% aprotonin, 2.5 mMphosphate buffer at pH 7.0). The medium was decanted and discarded between incubations. At the conclusion of these incubations, the tissue fragments had swollen to 150% of their original volume. The tissue was then placed in a smooth-wall glass homogenizer and homogenized with five strokes of a Teflon pestle. The ionic strength was increased to approximately 100 mM by adding 1 ml 1.O M NaCl to the homogenate. After thorough mixing, the homogenate was centrifuged 2 min at 1500 rpm. The neurofilament-rich supernatant was removed and the neurofilaments were precipitated by centrifuging 10 min in an Airfuge batch rotor. The pellet was washed twice by bringing it to a volume of 3 ml with 0.1 M NaCl and centrifuging 10 min in the Airfuge. The pellet was finally resuspended in I..0 ml of the hypotonic medium and stored at -20°C. ParaJin Sections. Normal rats were anesthetized with chloral hydrate (400 mg/kg, i.p.) and perfused through the heart with Bouin’s fixative. The spinal

AXONAL

REGENERATION

87

cord was removed, dehydrated, embedded in paraffin, and sectioned at 10 pm thickness. Individual sections were cut from the ribbon and deparaffinized in hexane. The tissue sections were hydrated in ascending ethanol series and one, two, or three sections were placed in a microtiter well. The ELISA assay was carried out as above. Sections adjacent to these were mounted on glass slides and stained with hematoxylin and eosin to verify the quality of fixation and embedding. Surgical Procedures.Female Wistar rats weighing 200 to 250 g were anesthetized with chloral hydrate (400 mg/kg, i.p.) and prepared for sterile surgery by shaving the hair from the hindquarters and scrubbing the skin with 0.13% benzalkonium chloride (Zephiran). Surgical instruments were soaked 15 min in Zephiran prior to operating. An incision 2 cm long was made over the sciatic notch on the right side, and after retracting the skin and muscle, the sciatic nerve was crushed or transected 1 cm distal to the sciatic notch. Crushing was carried out with a nonserrated mosquito hemostat for 1 s. Transection was accomplished with iridectomy scissors, after which the nerve stumps were ligated with 6-O silk thread and separated by 1 .O cm to prevent regeneration from occurring. The overlying muscle was sutured with interrupted 4-O silk and the skin was stapled with wound clips. At intervals of 0 to 16 weeks postoperatively, the animals were reanesthetized, and the entire sciatic nerve was removed, stretched to resting length on a piece of cardboard, and frozen on dry ice. Five single-edge razor blades were mounted in a specially fashioned plastic holder which firmly held the blades precisely 3 mm apart. The center blade was placed over the lesion site and the nerve was sliced simultaneously into four 3-mm-long segments, thus providing two specimens proximal to the lesion and two specimens distal to the lesion. Control unoperated specimens were obtained in an identical manner from the contralateral limb of each experimental rat. Each specimen was homogenized in 3 ml homogenizing medium and stored at -20°C until assayed by the ELISA procedure. RESULTS

Characteristics of the Fluorometric ELISA Assay The Alkaline PhosphataseELBA Reaction. The alkaline phosphatase reaction was linear over a 40-min incubation duration (Fig. 1) at all but the highest concentrations of homogenate (500 ng protein per well). In addition, the reaction rate (i.e., the slope of each line) was directly proportional to the amount of protein in the wells. This relationship is illustrated in Fig. 2. Linear regression was used to obtain the slope for each of the curves of Fig. 1. These values, which are a measure of the reaction rate, have been plotted in Fig. 2 as a function of tissue concentration. The resulting linear relationship

88

GUTH

0

IO

20 Incubation

ET AL.

Duration

(minutes)

FIG. 1. The quantitative ELBA assay for phosphotylated neurofilament (PNF) was plotted as a function of incubation duration to illustrate that the reaction rate is linear over a wide range of substrate concentrations.

indicates that in this ELBA assay procedure, the slope can be used to determine the concentration of PNF in a tissue homogenate. Furthermore, the procedure is sufficiently sensitive to measure PNF readily in 50 ng of CNS protein. The protein content of CNS tissues is approximately 10% (w/v), consequently the minimum amount of tissue required for measuring PNF

.-6 $

1.8-

",

1.4-

-z

l.Z-

1.6-

2 l.O'C 2 1.8a ; 0.6g

0.4-

5

0.20.

0

.

, loo

,

, 200 Protein

300 (rig/well)

400

500

FIG. 2. The rate of the alkaline phosphatase ELBA reaction (dY/dT of Fig. 1) when plotted as a function ofprotein concentration is directly proportional to the PNF content ofthe homogenate.

AXONAL

89

REGENERATION

Protein (pg/well) FIG. 3. ELBA assayof a purified preparation of PNF showed that the reaction rate was directly proportional to the concentration of PNF. Purified preparations of PNF can therefore be used to determine the absolute amount of PNF in a tissue homogenate.

content is 500 ng wet weight/50 ~1, which is equivalent to a 1: 100,000 (w/v) homogenate. Assay of Purified Phosphorylated Neurojlaments. To develop a standard of comparison for experimental studies, PNF protein was purified from rat brain (see Materials and Methods) and assayed by the ELISA procedure (Fig. 3). Comparison of the results obtained with this purified PNF preparation with that obtained from the crude homogenate (Fig. 2) revealed that 500 pg of purified PNF gave approximately the same reaction rate (0.70) as did 250 ng of brain homogenate. This result indicates that the purification procedure had concentrated the PNF approximately 500-fold. Measurement of Phosphorylated Neurofilament from Parafin Sections of Rat Spinal Cord. The wet weight of a single lo-pm section of rat spinal cord is well within the limits of sensitivity of the assay. We placed one, two, or three deparaffinized and rehydrated transverse sections of spinal cord in each well of a tray. The tissue samples were dried and the ELISA assay was carried out. The results displayed in Fig. 4 show that the reaction was directly proportional to the number of sections in the well and that the method was sensitive enough for measurement of PNF in single sections of spinal cord that had been fixed and embedded in paraffin. Application

of the ELBA Assay to the Quantitative of Nerve Degeneration and Regeneration

Analysis

Proximodistal Distribution of Phosphorylated Neuro_filament along the Normal Sciatic Nerve. Because nerve branches emerge from the sciatic nerve

90

GUTH ET AL.

No. of Sections per Well FIG. 4. Transverse paraffin sections of rat spinal cord were individually deparaffinized, rehydrated, and assayed for PNF content. The PNF content was directly proportional to the number of tissue sections per well, showing that the assaycan be applied to histologic preparations.

at frequent intervals, the number of axons and the PNF content of the nerve might be expected to show a proximodistal decline. To evaluate this possibility, we examined the PNF content of unoperated control nerves taken from the contralateral side of the experimental animals. We combined the results from all control nerves (regardless of whether they were obtained from rats that had been operated 1, 2, 3,4, or 8 weeks previously) because we found no consistent changes in their PNF content attributable to postoperative duration. However, we did find a significant proximodistal decline in PNF content along successive 3-mm segments of the unoperated sciatic nerve as nerve branches left the main trunk. Counting PNF content of the most proximal specimen as lOO%, the succeeding three specimens were 90.6, 87.9, and 59.9%, respectively. Consequently, in experiments involving measurement of PNF content of experimental and control sciatic nerves, we were careful to match the levels from which the experimental and control specimens were chosen. Changes in Phosphorylated NeuroJilament Content Distal to the Site of Nerve Crush or Transection. For this analysis we pooled the two 3-mm samples distal to the nerve lesion and compared them with the comparable levels on the contralateral unoperated side. The data were normalized by expressing the results on the operated side as a percentage of those on the unoperated side (Fig. 5 and Table 1). Note that when regeneration into the distal stump was prevented (e.g., after nerve transection), the PNF content of the distal stump declined gradually and reached a very low level by 4 weeks postoperatively. However, when reinnervation of the distal stump by regenerating

AXONAL

91

REGENERATION

Weeks

Postoperative

FIG. 5. Changes in PNF content distal to crush (closed circles) and transection (open circles) lesions of rat sciatic nerve. After either procedure the PNF content declined initially; it subsequently returned toward normal if regeneration was permitted (crush lesion) but not if regeneration was prevented (transection lesion).

nerve fibers was permitted (e.g., after nerve crush), the PNF content of the distal stump showed a lesser rate of initial decrease and subsequently returned to 60% of control value by 8 weeks postoperatively.

Use of the ELISA Phosphorylated Neurofilament Assayas a Quantitative Measure of Regeneration. The data in Fig. 5 show that after sciatic nerve transection, a measurable concentration of PNF remained in the distal nerve stump for 3 weeks. This relatively slow decline in PNF presumably reflects TABLE 1 Comparison of Changes in Phosphorylated Neurofilament Content in Regions Immediately Proximal and Distal to Crush and Transection Injuries of the Rat Sciatic Nerve’ Crush lesion

Transection lesion

Weeks postop

No.

Proximal

Distal

No.

Proximal

Distal

1 2 3 4 8 16

4 5 4 3 5 4

112 88 99 34 76 80

73 42 42 14 59 88

4 3 6 3 5 4

117 164 168 5-l 64 31

53 16 10 4 6 0

’ Values are percentages of respective controls.

92

GUTH

ET AL.

Weeks Postoperative FIG. 6. Sequential changes in PNF of the distal segment of a crushed sciatic nerve after correction for the PNF content of residual degenerating nerve fibers. This figure provides an estimate of the regeneration of axons into the distal stump.

the gradual nature of axonal degeneration and the relatively slow removal of the products of degeneration by macrophages. After sciatic nerve crushing, however, the PNF concentration was higher at each of these early postoperative intervals, presumably because the distal stump was being reinnervated by regenerating axons. Consequently, to obtain a meaningful quantitative picture of the rate of axonal ingrowth, the PNF content attributable to degenerating nerve fibers would have to be subtracted; we have done this in Fig. 6 by plotting the difference in PNF concentration of the crushed vs the transected nerves at each postoperative time interval. For this purpose we used the slope measurements directly without correcting for control values. Consequently the values depicted in this figure are expressed in units different from those in Fig. 5. The salient findings in Fig. 6 are (i) there was a delay of 0.5 weeks before the PNF content of the distal stump began to increase, and (ii) the rate of increase in PNF content was more rapid in the early postoperative stages than in the later ones.

Changes in Phosphorylated Neurofilament Immediately Proximal to a Nerve Lesion. Changes in PNF content were not restricted to the distal nerve stump, but also occurred in the region immediately proximal to a lesion of the sciatic nerve (Fig. 7 and Table 1). The data in Fig. 7 were obtained by combining the PNF content from the two 3-mm samples proximal to the lesion, expressing them as a percentage of the contralateral control, and plotting them as a function of postoperative duration. When the nerve was crushed to permit the injured axons to regenerate, the PNF content of the proximal region declined rather gradually to about 80% of normal. But when

AXONAL

93

REGENERATION

0 I234

8 Weeks

16

Postoperative

FIG. 7. Changes in PNF content proximal to crush (closed circles) and transection (open circles) lesions of rat sciatic nerve. Note that the PNF content increased during the 1st 3 weeks when regeneration was prevented (transection lesion) but not if regeneration was permitted (crush lesion).

the nerve was transected and ligated to prevent regeneration of the injured axons, there was an increase in PNF to 170% of control by the third postoperative week followed by a subsequent decline to subnormal values. DISCUSSION

Rationale for Selecting the Fluorometric Assay. We initially attempted a calorimetric ELISA method using a peroxidase-conjugated second antibody and hydrogen peroxide plus 2,2’-azinodi[3-ethyl-benzthiazoline sulfonate] (6) (ATBS, Boehringer-Mannheim, Inc.) or o-phenylenediamine (6) as substrate. Two deficiencies were noted with this approach: (i) Color development was maximal within 5 to 10 min, and the reaction had to be terminated within this brief interval in order for the optical density to be proportional to the concentration of enzyme. (ii) The calorimetric reaction was too insensitive to measure the low concentration of PNF in tissue homogenates. These problems were obviated by using the fluorometric phosphatase assay with methylumbelliferyl phosphate as substrate. In applying it to the ELISA procedure preliminary experiments were conducted to select optimal conditions for the assay (viz., concentration of antigen and antibodies, blocking reagent, and composition and pH of substrate). As a result, we obtained a simple and highly sensitive quantitative procedure. Utility of the Method. (i) Although the procedure was validated for quantitative measurement of PNF, it should be applicable to the determination of

94

GUTH

ET AL.

the concentration of any antigen for which a monoclonal antibody is available. Of course, to quantify the antigen concentration in absolute units, one must relate the content of the unknown to values obtained from a purified antigen standard. But even if purified antigen is not available, the linearity of the method enables one to evaluate relative changes under various experimental conditions. (ii) A second general feature of the method is its applicability to the quantitative assay of antigen in fixed, paraffin-embedded tissues. Although the rat spinal cord varies considerably in diameter along its longitudinal axis, the protein content of any lo-pm section of rat spinal cord is in the microgram range. The low fluorescence readings we obtained indicate that the fixation and paraffin-embedding may have resulted in significant loss of antigenicity. Sternberger (26) has emphasized that “different antigens vary in their susceptibility to fixation,” and he has recommended a systematic trial-and-error strategy in choosing fixatives for immunocytochemistry (26). By enabling us to measure antigen content in histologic tissue sections, the present quantitative ELISA method provides a more objective approach to the selection of fixatives for immunocytochemical studies. Application to Studies of Nerve Regeneration. Our experiments on sciatic nerve injury show that peripheral nerve regeneration can be assessed quantitatively by measuring changes in PNF content distal to the lesion. When regenerating neurites invade the distal stump segment of the crushed sciatic nerve, the PNF content of the distal stump increases demonstrably within 1 week and it undergoes a greater than 20-fold increase in the ensuing 8 weeks. On the other hand, when nerve regeneration is prevented, the PNF content of the distal stump declines markedly within the first postoperative week and reaches very low values within 4 weeks. These changes are consistent with immunocytochemical changes described in degenerating rat sciatic nerve (28). The increased PNF of the distal stump that occurs during axonal ingrowth represents two discrete processes, viz., the initial regenerative ingrowth of the axons and the subsequent maturation (increase in diameter) of the regenerated axons. There is indeed direct experimental evidence that anterograde transport of neurofilaments plays an important role in regulating the diameter of regenerating axons (13). To distinguish the relative contribution of axonal growth and axonal maturation to the observed changes in PNF content would require correlation of quantitative ELISA measurements with quantitative ultrastructural ones, a subject beyond the scope of the present study. In the absence of such information, the increase in PNF during nerve regeneration should be considered simply as indicative of increased content of axoplasm in the distal stump. Comparison of Changes in Phosphorylated Neurojilament Proximal and Distal to a Nerve Lesion. An unanticipated finding was the different nature

AXONAL

REGENERATION

95

of the response in the proximal and distal nerve stumps. Table 1 is presented to facilitate comparison of the changes in PNF content in the regions immediately proximal and distal to crush and transection lesions. When regeneration was permitted (as after a crush lesion) there was a rather small and gradual decline in PNF content proximal to the site of injury (Table 1 and Fig. 6) and a gradual increase in the distal nerve stump (Table 1 and Fig. 7). But when regeneration was not permitted (as after the transection lesion) there was a significant and steady accumulation of PNF in the proximal stump during the first 3 postoperative weeks and a subsequent decline thereafter (Table 1 and Fig. 6), accompanied by a steady decline in PNF content of the denervated distal nerve stump. The damming of axoplasm proximal to a nerve lesion is indicative of proximodistal movement of axoplasm (3 l), and the accumulation of specific proteins in such axoplasmic swellings has been considered as evidence for the intraaxonal transport of these macromolecules (8). Consequently, the accumulation of PNF proximal to injury in which regeneration is prevented, and not after injury in which regeneration is permitted, is consistent with the hypothesis that PNF is axonally transported (12, 13,22). The de n~vo appearance of PNF in cell bodies of regenerating neurons (20) is also consistent with the concept that PNF is synthesized in the cell body and is transported to the axon. CONCLUSION The method described in this contribution, having been validated for the quantification of peripheral nerve regeneration, should be applicable in the following ways to investigations on the central nervous system. (i) It will permit quantitative evaluation of experiments on axonal development, neuronal plasticity, neuritic sprouting, and nerve fiber regeneration in the brain and spinal cord. (ii) It can provide quantitative information for assessing the success of brain grafts and of growth of CNS axons into peripheral nerve grafts. (iii) It will prove helpful in the statistical evaluation of treatments designed to promote regenerative growth of nerve fibers into the injured CNS. REFERENCES 1. AGUAYO, A. J. 1985. Axonal regeneration from injured neurons in the adult mammalian central nervous system. Pages 457-484 in C. W. COTMAN, Ed., Synaptic Plasticity and Remodeling Guilford Press, New York. 2. BJORKLUND, A., AND U. STENEVI. 1984. Intracerebral neural implants: neuronal replacement and reconstruction of damaged circuitries. Annu. Rev. Neurosci. 7: 279-308. 3. CARDEN, M. J., W. W. SCHLAEPFER, AND V. M.-Y. LEE. 1985. The structure, biochemical properties, and immunogenicity of neurofilament peripheral regions are determined by the phosphorylation state. J. Biol. Chem. 260: 9805-98 17.

96

GUTH

ET AL.

4. DAHL, D., AND A. BIGNAMI. 1986. Neurofilament phosphorylation in development: a sign of axonal maturation? Exp. Cell Res. 162:220-230. 5. DE LA TORRE, J. C. 1981. Spinal cord injury: review of basic and applied research. Spine 6: 315-335. 6. DOHERTY, P., J. D. DICKSON, T. P. FLANIGAN, AND F. S. WALSH. 1984. Quantitative evaluation of neurite outgrowth in cultures of human foetal brain and dorsal root ganglion cells using an enzyme-linked immunoadsorbent assay for human neurofilament protein. J. Neurochem. 42: 1116-l 122. 7. FADEN, A. I., T. P. JACOBS, M. T. SMITH, AND J. W. HOLADAY. 1983. Comparison of thyrotropin-releasing hormone (TRH), naloxone, and dexamethasone treatments in experimental spinal injury. Neurology33: 673-678. 8. GRAFSTEIN, B., AND D. S. FORMAN. 1980. Intracellular transport in neurons. Physiol. Rev. 60: 1167-1283. 9. GUTH, L., C. P. BARREN, E. J. DONATI, F. D. ANDERSON, M. V. SMITH, AND M. LIFSON. 1985. Essentiality of a specific cellular terrain for growth of axons into a spinal cord lesion. Exp. Neurol. 88: 1- 12. 10. GUTH, L., C. P. BARRETT, E. J. DONATI, M. V. SMITH, M. LIFSON, AND E. ROBERTS. 1985. Enhancement of axonal growth into a spinal cord lesion by topical application of triethanolamine and cytosine arabinoside. Exp. Neural. 88: 44-55. 11. HALL, E. D., AND J. M. BRAUGHLER. 1982. Glucocorticoid mechanisms in acute spinal cord injury: a review and therapeutic rationale. Surg. Neural. 18:320-327. 12. HOFFMAN, P. N., AND R. J. LASEK. 1975. The slow component of axonal transport. Identification of major structural polypeptides ofthe axon and their generality among mammalian neurons. J. Cell. Biol. 66: 35 l-366. 13. HOFFMAN, P. N., G. W. THOMPSON, J. W. GRIFFIN, AND D. L. PRICE. 1985. Changes in neurofilament transport coincide temporally with alterations in the caliber of axons in regenerating motor fibers. J. Cell Biol. 101:1332- 1340. 14. JULIEN, J.-P., AND W. F. MUSHYNSKI. 1983. The distribution of phosphorylation sites among identified proteolytic fragments of mammalian neurofilaments. J. Biol. Chem. 258: 4019-4025. 15. LUNDBORG, G., L. B. DAHLIN, N. DANIELSEN, R. H. GELBERMAN, F. M. LONGO, H. C. POWELL, AND S. VARON. 1982. Nerve regeneration in silicone chambers: influence of gap length and of distal stump components. Exp. Neurol. 76: 361-375. 16. MADISON, R. D., C. DA SILVA, P. DIKKES, R. L. SIDMAN, ANDT.-H. CHIU. 1987. Peripheral nerve regeneration with entubulation repair: comparison of biodegradable nerve guides versus polyethylene tubes and the effects of a laminin-containing gel. Exp. Neural. 95: 378-390. 17. NIETO-SAMPEDRO, M., AND C. W. COTMAN. 1985. Growth factor induction and temporal order in CNS repair. Pages 407-445 in C. W. COTMAN, Ed., Synaptic Plasticity and Remodeling Guilford Press, New York. 18. REDINBAUGH, M. G., AND R. B. TURLEY. 1986. Adaptation ofthe bicinochinicacid protein assay for use with microtiter plates and sucrose gradient fractions. Anal. Biochem. 153: 267-271. 19. REIER, P. J., AND J. D. HOULE. 1987. The glial scar: its bearing upon axonal elongation and transplantation approaches to CNS repair. Adv. Neurol., in press. 20. ROSENFELD,J., M. E. DORMAN, J. W. GRIFFIN, L. A. STERNBERGER,N. H. STERNBERGER, B. G. GOLD, AND D. L. PRICE. 1987. Distribution of neurofilament antigens after axonal injury. J. Neuropathol. Exp. Neurol. 46: 269-282. 2 1. SCHLAEPFER,W. W. 1977. Studies on the isolation and substructure of mammalian neurofilaments. J. Ultrastruct. Res. 61: 149-157.

AXONAL

REGENERATION

97

22. SZHLAEPFER, W. W. 1987. Neurofilaments: structure, metabolism and implications in disease. J. Neuropathol. Exp. Neural. 46: 117- 129. 23. SCHLAEPFER, W. W., AND L. A. FREEMAN. 1978. Neurofilament proteins of rat peripheral nerve and spinal cord. J. Cell Biol. 78: 653-662. 24. SISKEN, B. F., E. ROBERTS, AND I. G~ETZ. 1985. Triethanolamine, Tris, Hepes, and cytosine arabinoside show neuritogenic activity in cultured embryo ganglia. Exp. Neural. 88: 27-43. 25. SMITH, P. K., R. I. KROHN, G. T. HERMANSON, A. K. MALLIA, F. H. GARTNER, M. D. PROVENZANO, E. K. FUJIMOTO, N. M. GOEKE, B. J. OLSON, AND D. C. KLENK. 1985. Measurement of protein using bicinochinic acid. Anal. Biochem. 150: 76-85. 26. STERNBERGER,L. A. 1986. Pages 2 12-2 18 in Zmmunocytochemistry, 3rd Ed. Wiley, New York. 27. STERNBERGER,L. A., AND N. H. STERNBERGER. 1983. Monoclonal antibodies distinguish phosphorylated and nonphosphorylated forms of neurofilaments in situ. Proc. Nat/. Acad. Sci. USA 80: 6 126-6 130. 28. TROJANOWSKI, J. Q., V. M.-Y. LEE, AND W. W. SCHLAEPFER. 1984. Neurofilament breakdown products in degenerating rat and human peripheral nerves. Ann. Neurol. 16: 349355. 29. VARON, S., M. MANTHORPE, F. M. LONGO, AND L. R. WILLIAMS. 1983. Growth factors in regeneration of neural tissues. Pages 127-155 in F. J. SEIL, Ed., “Nerve, Organ, and Tissue Regeneration: Research Perspectives” Academic Press, New York. 30. VARON, S., M. MANTHORPE, AND L. R. WILLIAMS. 1984. Neuronotrophic and neurite promoting factors and their clinical potential. Dev. Neurosci. 6: 73- 100. 31. WEISS, P., AND H. B. HISCOE. 1948. Experiments on the mechanism of nerve growth. J. Exp. Zool. 107: 3 15-395. 32. WILLIAMS, L. R., F. M. LONGO, H. C. POWELL, G. LUNDBORG, AND S. VARON. 1983. Spatial-temporal progress of peripheral nerve regeneration within a silicone chamber. J. Comp. Neural. 218: 460-470.