Regular artilce
Ex vivo MR determined apparent diffusion coefficients correlate with motor recovery mediated by intraspinal transplants of fibroblasts genetically modified to express BDNF

https://doi.org/10.1016/S0014-4886(03)00036-0Get rights and content

Abstract

The purpose of this study was to determine whether apparent diffusion coefficients (ADCs) in ex vivo spinal cord white matter, calculated from diffusion weighted MR (DWI) images, correlate with axonal growth and behavioral recovery following subtotal hemisection and transplantation of fibroblasts genetically modified to express brain derived neurotrophic factor (BDNF). These genetically modified fibroblasts have been shown to promote axonal growth, diminish retrograde degenerative changes in axotomized Red nucleus neurons, and are associated with behavioral recovery. Since changes in ADC appear to reflect damage to axons and myelin sheaths, which conventional MR techniques do not identify, partial repair mediated by BDNF-secreting fibroblasts should be detected with ADC measures. Accordingly, we transplanted unmodified fibroblasts (Fb-UM) or fibroblasts modified to secrete BDNF (Fb-BDNF) into cervical subtotal hemisection cavities in adult rats. Rats with Fb-BDNF transplants showed significantly greater behavioral recovery over 12 weeks, as measured by tests of forelimb exploration and open field locomotor activity. Lesion sizes and transplant survival did not differ between the two groups, but immunocytochemical examination showed substantial growth of axons into the Fb-BDNF grafts and little growth into the Fb-UM grafts. Fixed spinal cords were imaged in a 9.4-T magnet. ADCs perpendicular (tADC) and parallel (lADC) to the long axis of the cord were measured in the dorsal lateral white matter, rostral and caudal to the transplant. tADC values and anisotropy index (AI = tADC/lADC) were elevated in both transplant types, indicating white matter damage, but were closer to normal in rats with Fb-BDNF, consistent with known neuroprotection and axonal growth elicited by BDNF. Closer to normal tADC and AI values correlated with improved behavioral recovery. These findings suggest that high-resolution imaging with measurement of tADC and lADC can provide a measure of functionally significant repair that may otherwise go undetected with conventional MR techniques.

Introduction

Early administration of methylprednisolone, which acts by limiting the secondary injury caused by an immune response, is associated with some improvement in functional outcome in patients following acute spinal cord injury Bracken et al 1992, Schwab 1996. Studies of spinal cord repair, however, have identified additional interventions that support axonal growth and recovery of function. These advances will certainly be translated into the clinic in the near future Lu and Waite 1999, Murray 2001. Because histologic confirmation is not feasible in a clinical setting, and behavioral improvement may proceed more slowly in humans than in small animal models, it is essential to develop a noninvasive modality that can monitor response to treatment.

Currently, magnetic resonance imaging (MRI)1 is the best modality for evaluation of spinal cord parenchyma following injury, and correlation has been demonstrated between initial results with conventional MR and clinical outcome in humans Kulkarni et al 1987, Ramon et al 1997, Flanders et al 1999, Takhtani and Melhem 2000, Ditunno et al 2002. Follow-up imaging of spinal cord injury, however, has been limited to assessment of spinal cord morphology and development of posttraumatic syringomyelia Quencer et al 1986, Milhorat et al 1995, Schurch et al 1996, Jinkins et al 1998, Schwartz et al 1999a. Wallerian degeneration following injury is seen with conventional MR as changes in signal intensity above and below the site of injury (Becerra et al., 1995), but these changes cannot be differentiated from other pathologies such as edema. MR imaging in clinical studies of human embryonic spinal cord grafts placed in posttraumatic spinal cord cysts was also confined to postoperative follow-up of cyst obliteration by the grafts Falci et al 1997, Wirth et al 2001. Since conventional MRI appears limited to defining macroscopic changes in the spinal cord following injury, another technique is needed to evaluate microscopic histologic parameters that accompany potential treatments and are functionally significant, but that do not alter overall spinal cord morphology.

Diffusion weighted MRI (DWI) of the CNS can provide more information concerning cell structure and integrity than conventional MR techniques. DWI reflects water diffusion, on the scale of microns, in a specific direction. In neural tissue, the cell membranes and myelin sheaths of axons form barriers to water diffusion, and this property makes DWI well suited for evaluating perturbations in this system. Intact white matter tracts may be expected to display anisotropy (the tendency for water to diffuse preferentially in one direction) with increased diffusion parallel to the long axis of the axons compared to the transverse (perpendicular to the long axis) direction. This is because the cell membranes and myelin sheaths form diffusion barriers in the transverse direction but not in the longitudinal direction Doran and Bydder 1990, Hajnal et al 1991, Barkovich 2000. In the spinal cord, where most axons are oriented longitudinally, the diffusion of water is therefore preferentially in the longitudinal direction. DWI can provide separate measurements of water diffusion in the transverse direction (tADC) and the longitudinal direction (lADC), and ex vivo animal studies have confirmed anisotropic water diffusion in the spinal cord white matter. These studies have demonstrated that the tADC is significantly smaller than the lADC, thus indicating preferential water diffusion in the longitudinal direction Ford et al 1994, Ford et al 1994. It is the spinal cord white matter that principally determines the anisotropic properties of the spinal cord. Gray matter is not nearly as anisotropic because it lacks the uniform structural orientation of axons that characterizes white matter Ford et al 1994, Inglis et al 1997, Pattany et al 1997, Schwartz et al 1999b.

While it is not yet technically feasible to obtain the necessary resolution in vivo for evaluating small portions of the spinal cord with DWI, ex vivo studies have shown that DWI is more sensitive than conventional MRI techniques in evaluating the extent of spinal cord injury. Changes in ADC values measured by DWI identified areas of spinal cord injury surrounding a contusion site that otherwise appeared normal with conventional MRI techniques (Ford et al., 1994). DWI and the measurement of ADC values are also more sensitive than conventional MRI in the early detection of spinal cord cavities following excitotoxic injury (Schwartz et al., 1999b). Finally, it has recently been shown that an anisotropy measure, based on ADC values obtained with DWI, could detect the effects of a neuroprotective agent following a spinal cord contusion injury in adult rats; this anisotropy measure also appear to correspond with behavioral outcome (Nevo et al., 2001).

In this study, we transplanted fibroblasts genetically modified to secrete brain derived neurotrophic factor (BDNF) into a cervical lateral funiculus lesion to determine whether the measurement of apparent diffusion coefficients in the surrounding white matter would correlate with behavioral recovery. BDNF acts on the TrkB receptors located on rubrospinal tract (RST) axons and Red nucleus neurons, and BDNF secreting fibroblasts placed into spinal cord lesions have been shown to promote neuroprotection of Red nucleus neurons, regeneration of RST axons, and growth of other axons. Animals with BDNF secreting fibroblasts also demonstrated improved behavioral recovery compared with animals receiving transplants of unmodified fibroblasts. Kim et al 1999, Kim et al 2001, Lu and Waite 1999, Himes and Tessler 2001. We measured the tADC and lADC in the lateral white matter of the spinal cord rostral and caudal to genetically modified or unmodified grafts following a lateral funiculus lesion. ADC measurements were also obtained in the cervical white matter of normal controls. As a measure of anisotropy, we used the ratio tADC/lADC, with lower values indicating increased anisotropy and preferential water diffusion in the longitudinal direction. Similar measurements were also taken within the fibroblast grafts to determine whether the ingrowth of axons could be detected. Lesion extent was measured by calculating the area of spared right-sided white and gray matter in the spinal cord at the midpoint of the graft.

Section snippets

Subjects

Eighteen female Sprague–Dawley rats (250–300 g; Taconic, Germantown, NY) were studied. Animals were divided into three groups. One group (n = 6) received a lateral funiculus lesion on the right side of the spinal cord and a transplant of fibroblasts that had been genetically modified to secrete BDNF (Fb-BDNF). An operated control group (n = 8) received the same surgery with a transplant of unmodified fibroblasts (Fb-UM), which do not secrete BDNF. All operated animals underwent preoperative

Histological analysis: graft and lesion extent

All grafts survived and were closely apposed to the host spinal cord. Few cysts developed. The host–graft interface is readily distinguishable in Nissl myelin-stained sections because the grafted fibroblasts are small, densely packed, and darkly stained. All lesions ablated the dorsal lateral funiculus. There was, however, considerable variability in lesion size and many approached a complete hemisection. The average remaining white and gray matter at the lesion center, measured from histologic

Discussion

We show that the measurement of apparent diffusion coefficients in the spinal cord white matter can assess the effects of a previously validated model of treatment for spinal cord injury that is neuroprotective, stimulates axonal growth, and promotes behavioral recovery Liu et al 1999, Kim et al 2001. The white matter rostral and caudal to BDNF-secreting fibroblast grafts had tADC and AI values closer to normal than white matter surrounding unmodified fibroblasts. There were no differences seen

Acknowledgements

Funding for this research has been provided by NIH grants NS02230, NS24707, NS25921, the Research Services of the Department of Veteran Affairs, and American Society of Neuroradiology/Berlex Basic Science Fellowship. The authors thank the following people for their help and advice: Theresa Connors, Lisa Hodge, Maureen Tumolo, Chris A. Tobias, B. Timothy Himes, Steve Han, Stacy Jacob, and Carla Tyler-Polsz.

References (69)

  • E.D. Schwartz et al.

    Diffusion-weighted imaging of the spinal cord

    Neuroimaging Clin. N. Am.

    (2002)
  • E.D. Wirth et al.

    Dynamic assessment of intraspinal neural graft survival using magnetic resonance imaging

    Exp. Neurol.

    (1995)
  • A.J. Barkovich

    Concepts of myelin and myelination in neuroradiology

    Am. J. Neuroradiol.

    (2000)
  • D.M. Basso et al.

    A sensitive and reliable locomotor rating scale for open field testing in rats

    J. Neurotrauma

    (1995)
  • J.L. Becerra et al.

    MR-pathologic comparisons of Wallerian degeneration in spinal cord injury

    Am. J. Neuroradiol.

    (1995)
  • M. Bilgen et al.

    In vivo magnetic resonance microscopy of rat spinal cord at 7 T using implantable RF coils

    Magn. Reson. Med.

    (2001)
  • M.B. Bracken et al.

    Methylprednisolone or naloxone treatment after acute spinal cord injury1-year follow-up data: results of the second National Acute Spinal Cord Injury Study

    J. Neurosurg.

    (1992)
  • D.L. Buckley et al.

    The effect of ouabain on water diffusion in the rat hippocampal slice measured by high resolution NMR imaging

    Magn. Reson. Med.

    (1999)
  • C.L. Chin et al.

    Biexponential diffusion attenuation in the rat spinal cordcomputer simulations based on anatomic images of axonal architecture

    Magn. Reson. Med.

    (2002)
  • J.F. Ditunno et al.

    Predicing outcome in traumatic spinal cord injury

  • M. Doran et al.

    Magnetic resonanceperfusion and diffusion imaging

    Neuroradiology

    (1990)
  • S. Falci et al.

    Obliteration of a postraumatic spinal cord cyst with solid human embryonic spinal cord graftsfirst clinical attempt

    J. Neurotrauma

    (1997)
  • D.A. Fenyes et al.

    In vivo diffusion tensor imaging of rat spinal cord with echo planar imaging

    Magn. Reson. Med.

    (1999)
  • A.E. Flanders et al.

    The relationship between the functional abilities of patients with cervical spinal cord injury and the severity of damage revealed by MR imaging

    Am. J. Neuroradiol.

    (1999)
  • J.C. Ford et al.

    Numerical model for calculation of apparent diffusion coefficients (ADC) in permeable cylinderscomparison with measured ADC in spinal cord white matter

    Magn. Reson. Med.

    (1997)
  • J.C. Ford et al.

    MRI characterization of diffusion coefficients in a rat spinal cord injury model

    Magn. Reson. Med.

    (1994)
  • J.C. Ford et al.

    Dependence of apparent diffusion coefficients on axonal spacing, membrane permeability, and diffusion time in spinal cord white matter

    Magn. Reson. Imaging

    (1998)
  • F. Franconi et al.

    In vivo quantitative microimaging of rat spinal cord at 7T

    Magn. Reson. Med.

    (2000)
  • S.C. Grant et al.

    MR microscopy of multicomponent diffusion in single neurons

    Magn. Reson. Med.

    (2001)
  • V. Gulani et al.

    A multiple echo pulse sequence for diffusion tensor imaging and its application in excised rat spinal cords

    Magn. Reson. Med.

    (1997)
  • V. Gulani et al.

    Apparent diffusion tensor measurements in myelin-deficient rat spinal cords

    Magn. Reson. Med.

    (2001)
  • J.V. Hajnal et al.

    MR imaging of anisotropically restricted diffusion of water in the nervous systemtechnical, anatomic, and pathologic considerations

    J. Comput. Assisted Tomogr.

    (1991)
  • B.T. Himes et al.

    Neuroprotection from cell death following axotomy

  • S. Hwang et al.

    The effect of myelin loss on the diffusion-sensitized MRsignal in the spinal cordapplication of a new finite difference method

    Radiology

    (2000)
  • Cited by (0)

    View full text