Diffusion tensor imaging findings suggestive of white matter alterations in a canine model of mucopolysaccharidosis type I

Abstract

Purpose

We investigated fractional anisotropy (FA) and radial diffusivity (RD) in a canine model of mucopolysaccharidosis (MPS). We hypothesized that canines affected with MPS would exhibit decreased FA and increased RD values when compared to unaffected canines, a trend that has been previously described in humans with white matter diseases.

Methods

Four unaffected canines and two canines with MPS were euthanized at 18 weeks of age. Their brains were imaged using high-resolution diffusion tensor imaging (DTI) on a 7T small-animal magnetic resonance imaging system. One hundred regions of interest (ROIs) were placed in each of four white matter regions: anterior and posterior regions of the internal capsule (AIC and PIC, respectively) and anterior and posterior regions of the centrum semiovale (ACS and PCS, respectively). For each specimen, average FA and RD values and associated 95% confidence intervals were calculated from 100 ROIs for each brain region.

Results

For each brain region, the FA values in MPS brains were consistently lower than in unaffected dogs, and the RD values in MPS dogs were consistently higher, supporting our hypothesis. The confidence intervals for affected and unaffected canines did not overlap in any brain region.

Conclusion

FA and RD values followed the predicted trend in canines affected with MPS, a trend that has been described in humans with lysosomal storage and dysmyelinating diseases. These findings suggest that the canine model parallels MPS in humans, and further indicates that quantitative DTI analysis of such animals may be suitable for future study of disease progression and therapeutic response in MPS.

Introduction

The domestic dog, Canis familiaris, suffers from almost 360 diseases found in humans.¹ Examples include colonies of canines affected with lysosomal storage disorders (LSDs) such as Krabbe disease,² GM1 gangliosidosis,³ and mucopolysaccharidosis type I (MPS I).⁴ MPS causes disability and early death by a mutation in α-L-iduronidase, an enzyme required for the intra-lysosomal breakdown of glycosaminoglycans (GAGs). The breakdown products of partially degraded GAGs accumulate in lysosomes, thereby producing cellular dysfunction in almost every organ system, including the brain where disease affects cerebral white matter.

Because LSDs are rare in humans, establishment of a valid canine model of MPS I would provide researchers the opportunity to better understand disease pathogenesis and develop potential treatments. For example, one group has treated MPS I-affected canines with intrathecal enzyme replacement therapy and used neuroimaging findings to help guide future treatment in humans.⁵

In humans with MPS I, common imaging findings include hydrocephalus, atrophy, enlarged perivascular (or Virchow-Robin) spaces, and white matter changes.^6,7 Magnetic resonance (MR) imaging provides a noninvasive means to potentially study disease progression and therapeutic effect of treatments. In particular, diffusion tensor imaging (DTI) is a promising means to study the changes in white matter.

DTI is a type of MR imaging used to study the water diffusivity in tissues. In many white matter structures, myelin acts as a powerful diffusion barrier forcing diffusion to proceed in a primary direction along an axon, which increases anisotropy. Two DTI metrics commonly used in diffusion analysis include fractional anisotropy (FA) and radial diffusivity (RD). Both of these metrics are scalar values, where an FA value of one represents diffusion that is completely restricted to one direction, while RD represents diffusion along the two minor diffusion axes.

The purpose of this study was to determine whether the DTI findings in white matter of dogs affected by MPS I differ from that of unaffected dogs. We hypothesized that the white matter of canine brains affected with MPS would have decreased FA and increased RD values when compared to unaffected canines, as is seen in humans with lysosomal storage and dysmyelinating diseases.^6,8,9 If that were the case, then DTI might prove to be a reliable means to study canine MPS I brains in longitudinal studies for assessment of treatment response.

Methods

Specimen acquisition and preparation

Six post-mortem canine brains, two affected with MPS I and four unaffected, were obtained from Iowa State University Department of Animal Science. The canines had been euthanized at 18 weeks as part of an independent research study that had been performed under the auspices of the Institutional Animal Care and Use Committee (IACUC). All canines were administered intravenous (IV) heparin (0.5 ml, 500 U/dog) 10 minutes before euthanization with Euthasol (200 mg/kg), until respiration and cardiac activity had stopped and corneal and pedal reflexes were absent. Perfusion catheters were placed in each of the carotid arteries, the arteries clamped below the catheter placement and the abdominal aorta cut. The dogs were perfused first with a 0.9% sodium chloride solution and then with a solution of 10% Magnevist (Bayer Healthcare Pharmaceuticals, Whippany, NJ) in 10% formalin.

Following perfusion, most of the skin, subcutaneous tissue, and muscle were removed from the head. The heads were then placed in a 10% formalin solution and allowed to post-fix for 24 hours at room temperature. Then the heads were transferred to a solution of 1% Magnevist in 10% formalin in order to lower the T1 of brain tissue (which would allow for a shorter MR imaging time) and shipped to our institution at room temperature. Upon receipt of the heads, we carefully removed the brains from the skulls and placed them in a solution of 1% ProHance (Bracco Diagnostics Inc, Monroe Township, NJ) in 10% formalin. Two weeks before imaging, the brains were immersed in a solution of 1% ProHance in phosphate-buffered saline (pH 7.4) in order to allow the tissue to rehydrate and reduce the T1 relaxation time.

Imaging

All brains were scanned using an identical protocol on a 7T small-animal MR imaging system (Magnex Scientific, Yarnton, Oxford, England, UK) equipped with 670 mT/m Resonance Research gradient coils (Resonance Research Inc., Billerica, MA) using a 65 mm internal diameter quadrature radiofrequency (RF) coil (M2M Imaging, Cleveland, OH). The system was controlled with an Agilent Direct Drive console (Agilent Technologies, Santa Clara, CA). The images were acquired using a custom-designed six-direction diffusion-weighted spin-echo pulse sequence (repetition time (TR) = 100 ms, echo time (TE) = 18.1 ms, number of excitations (NEX) = 1, b value = 1506 s/mm²). The acquisition matrix was optimized to fit the dimensions of each canine brain and adjusted for a field of view producing a Nyquist-limited isotropic voxel size of 200 µm and a slice thickness of 200 µm. Diffusion preparation was accomplished using a modified Tanner-Stejskal diffusion-encoding scheme with a pair of unipolar, half-sine diffusion gradient waveforms. Total acquisition time was approximately 12 hours for each brain. Following image acquisition, the data for each brain were then smoothed using the SUSAN de-noising algorithm implemented in FSL with a three-voxel-kernel radius. For each brain, maps of the three eigenvalues, the three eigenvectors, and FA were reconstructed using Diffusion Toolkit Version 0.6.2. RD maps were calculated by averaging the λ₂ and λ₃ eigenvalue maps, which represent the two non-dominant diffusion axes as discussed above, using ImageJ Version 1.5. FA and RD are defined in terms of eigenvalues in equations 1 and 2, respectively.

$$\text{FA}=\sqrt{\frac{3}{2}}\frac{\sqrt{({\lambda }_1-{\lambda }_m{)}^2+({\lambda }_2-{\lambda }_m{)}^2+({\lambda }_3-{\lambda }_m{)}^2}}{\sqrt{{\lambda }_1^2+{\lambda }_2^2+{\lambda }_3^2}}$$ (1)

with ${\lambda }_m=\frac{{\lambda }_1+{\lambda }_2+{\lambda }_3}{3}$ or the mean of the eigenvalues, referred to as mean diffusivity

$$\text{RD}=\frac{{\lambda }_2+{\lambda }_3}{2}$$ (2)

DTI metric acquisition and comparison

We interrogated four brain regions: the anterior and posterior internal capsule (AIC and PIC respectively), and the anterior and posterior centrum semiovale (ACS and PCS respectively). To determine the FA and RD values within these brain regions, regions of interest (ROIs) were placed on coronal sections of the B₀ map for each brain using ImageJ. For each region, we defined the region’s anatomical boundaries in the dorsal, ventral, rostral, and caudal directions. We then selected 10 contiguous slices containing the most robust white matter volume within each of the four regions. These slices were defined using anatomical landmarks in order to ensure that comparable slices were sampled in all specimens and that the interrogation method would be reproducible. Once the boundaries were defined, two independent raters placed 10 non-overlapping, square ROIs (five per hemisphere) on each slice, for a total of 100 ROIs per brain region per specimen. For a complete description of ROI sampling method, refer to Li et al.¹⁰ These ROIs were then superimposed on the FA, RD, and the three eigenvalue maps, and average values of all five metrics within each ROI were measured.

For each of the six specimens studied, the mean FA and RD values were calculated for four brain regions, the AIC, PIC, ACS, and PCS, based on measurements from 100 ROIs per region. A 95% confidence interval was calculated and plotted for each mean FA and RD value in each of the six canines used in this study.

Results

In all four brain regions, a consistent pattern of significantly lower FA and higher RD values was seen in the MPS affected brains (Figure 1). For both FA and RD, the ACS and PCS showed the greatest difference between the cluster of measurements for the MPS canines and the cluster for the unaffected canines. Furthermore, no overlap was seen between the clusters of mean values for the two groups in any brain region.

Figure 1.

Average (a) fractional anisotropy and (b) radial diffusivity measurements for six specimens, four normal and two mucopolysaccharidosis-affected canines. For each of the four sampled brain regions, each data point represents an average of 100 regions of interest, with error bars indicating the 95% confidence interval for each mean.

When average FA and RD values were calculated for each brain region in both cohorts, the AIC had the highest average FA values both for the unaffected and MPS I specimens, 0.68 and 0.59, respectively; the ACS had the lowest average FA values of 0.63 for the unaffected specimens and 0.49 for the MPS I specimens. The highest average RD values were found in the PIC for the normal cohort and ACS for the MPS I cohort, 1.69 × 10⁻⁴ and 2.19 × 10⁻⁴, respectively. The lowest RD value for the normal cohort was measured in the PCS as 1.64 × 10⁻⁴ and in the PIC for the MPS I cohort as 2.07 × 10⁻⁴.

Discussion

In this study, we have shown that the white matter in MPS-affected canine brains has lower FA and higher RD values when compared to those of age-matched normal brains, consistent with our hypothesis. This study of ex vivo brains imaged over many hours with high-resolution DTI replicates the results we found using more conventional DTI performed in MPS I dogs in vivo.¹¹ In that study, we also found decreased FA values and elevated RD values; however, the ROIs were all placed in the corpus callosum. It was not possible to sample the corpus callosum reliably in the present study because of the small size of the corpus callosum in the relatively young dogs used.

The exact cause of these abnormal values in our MPS I cohort cannot be determined with certainty without histological analysis of the tissue that we imaged. Many factors can contribute to abnormally low FA values, including diminished myelination and axon loss. Abnormally high RD values have been found in mice lacking myelin.^12,13 From such data, investigators have postulated that the presence of normal degrees of myelination limit microscopic water diffusion perpendicular to the long axis of axons. Hence, it is possible that the abnormally high RD values seen in our MPS I cohort indicate diminished myelination in the areas of brain studied. Evidence suggesting that diminished myelination might contribute to the low FA values and elevated RD values in the MPS I cohort in this study can be found in a study performed by our group in vivo in dogs with MPS I.¹¹ In that study, we found that reduced FA and elevated RD values in the corpus callosum correlated with histological findings of abnormal composition of myelin and reduced expression of myelin-related genes.

Given that the ex vivo FA and RD changes in canines affected with MPS are similar to those found in vivo,^11,14 ex vivo imaging appears to show promise as a method for more definitive determination of alterations in the MPS brain. Furthermore, ex vivo imaging provides a means for more direct comparison between imaging findings and histological changes because images from MR scans can be co-registered with anatomical brain slices. Thus, a direct comparison of ROIs from imaging studies and histological slices can be performed. This process is especially important for DTI because so little is known about the histological underpinnings of DTI metrics.

A number of advantages are evident in the use of canine models of human diseases. First, canine brains are larger than murine brains, allowing for a more accurate evaluation of neurodegeneration on MR imaging.¹⁵ Second, the naturally occurring canine MPS model bears many similarities to the disease phenotype in humans, including valvular heart disease, plaques in large blood vessels, enlarged visceral organs, umbilical hernias, skeletal dysplasia, short stature, corneal clouding, and early death.^6,16

The fact that DTI changes in the brains of MPS I dogs showed marked changes in white matter appears to make the canine brain a suitable model for studying the human disease. This feature becomes especially important because data published by our group indicate that the canine model can serve as a suitable means for studying novel therapies for treating MPS I. In a previous study, we found that the DTI abnormalities in MPS I canines that were treated with enzyme replacement therapy were intermediate between unaffected carrier dogs and untreated MPS I dogs.¹¹ These findings suggested a treatment effect of the therapy and depict the manner in which DTI in canine models could potentially allow for assessment of such novel therapies.

Like any study, ours was subject to a number of limitations. One limitation is the small sample size, which was necessitated by the restricted availability of MPS I dogs. It will be important to determine whether our results can be reproduced with a larger sample size. Another limitation was imposed by the need to place ROIs by hand. This limitation was particularly evident in the diseased specimens, in which it was occasionally difficult to delineate the targeted brain regions because of white matter damage, or to avoid sampling some striations and Virchow-Robin spaces on the MPS brains. We also recognize some limitations of the use of the canine MPS I model of human disease. First, lysosomal storage diseases develop in humans as a result of many different mutations, but stem from a single mutation in canines. Second, untreated humans with MPS I rarely reach reproductive age and have marked hepatosplenomegaly, which is not the case in the canine model.¹⁶ Third, the onset of disease and rate of progression is predictable in the canine, but variable in humans.¹⁵ Nonetheless, the predictability of disease progression in affected dogs actually makes the canine model particularly useful in studying the effect of novel therapies, as well as in the investigation of novel uses of DTI data to evaluate disease, such as tensor shape analysis.¹⁷

In summary, we set out to determine whether the canine is a suitable MPS disease model for quantitative DTI analysis. We hypothesized that lower FA and higher RD values would be found in the white matter of MPS affected canines compared to unaffected canines. Our hypothesis was validated, and our findings are similar to those that have been previously described in humans with both lysosomal storage and dysmyelinating diseases.^6,8,9 These findings suggest that the canine model parallels MPS I in humans, and further indicates that quantitative DTI analysis of such animals may be suitable for future study of disease progression and therapeutic response in MPS I.

Funding

This work was performed at the Duke Center for In Vivo Microscopy, a national Biomedical Technology Resource Center supported by the National Institutes of Health/National Center for Research Resources/National Institutes of Biomedical Imaging and Bioengineering (grant P41 EB015897). This work was supported by Patricia Dickson’s grant: National Institutes of Health grant R01 NS085381, National Institute for Neurological Disorders and Stroke. James M. Provenzale and Steven Chen are paid consultants on this grant.

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.