Abstract
In response to a letter from Dr. Mark Butt (2010) expressing some concerns relating to our recent paper describing cytoplasmic vacuoles in cytons of dorsal root ganglion neurons in rats following prolonged exposure to neurotoxic organophosphates (Rogers-Cotrone et al. 2010), we would like to address his concerns in this Letter to the Editor.
The first point that we would like to address is Dr. Butt’s concern over the validity of our finding that these cytoplasmic vacuoles are related to the organophosphorus ester-induced delayed neurotoxicity that is seen following such long-term exposure to these compounds (Jortner et al. 2005). He feels that the vacuoles may represent an artifactual change or are simply spontaneous. As regards artifacts, he suggests that this effect may be caused by separation from adjacent cells, given the predilection for the vacuoles to occur in the periphery of the neuronal cell body. As indicated in our paper, we agree that the vacuoles originate in the periphery of the cyton, but they expand to involve more central regions of the cell body. In the course of our study, we examined hundreds of dorsal root ganglion neurons by transmission electron microscopy and failed to recognize artifactual separation from satellite cells, as demonstrated in Figure 4 of our paper. Given this information, we believe that there is little supportive evidence for cell separation being the cause of the vacuoles. Dr. Butt does not indicate other possible artifactual mechanisms, and a recent literature search did not reveal any published articles that describe artifact as the basis of the type of dorsal root ganglion neuronal vacuoles we described in our paper. We do not believe that the vacuoles were a result of tissue processing artifact, for the material we studied was well fixed and processed by accepted methods (Hancock et al. 2004). Thus, we respectfully disagree that the vacuoles are an artifactual change.
The second possibility that Dr. Butt raises is that the vacuoles represent a spontaneous (background) change. We do not disagree with this. In both the Introduction and Discussion sections of our paper, we have specifically indicated that this change can be seen at low levels in control animals, and we included references to support this contention. Thus, readers of our paper should not get the impression that vacuolation of dorsal root ganglion neurons cannot be seen in control animals.
Dr. Butt may be too restrictive in indicating artifact or background change as the sole bases of the vacuolation. There is a body of literature that indicates vacuoles in dorsal root ganglion neurons are also seen with injury to peripheral nerve branches of axons arising in these cells (Groves, Giometto, and Scaravilli 1997; Groves and Scaravilli 2005; Kerezoudi et al. 1995), and that the vacuoles are diminished following axonal regeneration (Groves, Giometto, and Scaravilli 1997). There is a reported increase in the vacuoles with age (Kerezoudi et al. 1995), a period when there is increase in peripheral nerve fiber degeneration (Kazui and Kohshiro 1988; Van Steenis and Kroes 1971). These data are relevant to our report. We have previously described ganglionic lesions in a rat model of organophosphorus ester-induced delayed neurotoxicity in which there is chemical injury to peripheral nerve branches arising from this population of neurons (Jortner et al. 2005). Indeed, the peripheral nerve lesions of this entity have been termed a “chemical transection” of the axon (Bouldin and Cavanagh 1979). As stated in our paper, we do not believe the vacuoles are a direct effect of exposure of dorsal root ganglion neurons to neurotoxic organophosphates but are, rather, a reflection of such toxicant-induced injury to the peripheral axonal branches and thus are relevant to the axonal injury studies noted above. We speculate that at least some of the vacuoles seen in control animals may relate to spontaneous, subclinical, axonal degeneration seen generally as a product of the age of the rat.
In response to Dr. Butt’s suggestion to “evaluate the appearance of these vacuoles with vacuoles known to develop as a degenerative change,” we offer the report of Kortz et al. (1997), who describe the presence of intraneuronal cytoplasmic vacuoles in young Rottweiler dogs with ataxia and weakness progressing to tetraparesis. At necropsy, these dogs had vacuoles that were essentially the same as the ones we described. Admittedly, the canine neuronal vacuoles were considerably more widespread, involving neurons in cerebral and cerebellar cortices, brainstem nuclei, spinal cord gray matter, peripheral autonomic ganglia, and dorsal root ganglia. In these cases there was an associated alteration of neuronal processes, including axonal degeneration, in cerebellar roof nuclei and regionally in the spinal cord white matter. We feel the Kurtz et al. study provides additional evidence that neuronal vacuoles qualitatively similar to what we described can be associated with neurologic disease.
Lastly, Dr. Butt expressed a concern with the method we have used to determine the proportion of affected cells in the ganglia, indicating that one must use stereological techniques (quantification in three dimensions of a tissue). Our observations were made from examination of photomicrographs of 1 µm thick toluidine blue and safranin stained sections through the ganglia and counted using the manual count tool of Nikon NIS-Elements Basic Research software. Admittedly not all of these details were indicated in the Methods section of our paper. Dr. Butt indicates that data derived from such an examination are flawed and urges Toxicologic Pathology to join “many other journals” in not accepting quantitative data from such two-dimensional 1 examination of tissue sections. Although Dr. Butt did not provide specific journal titles, we did examine recent issues of some pathology and neuropathology journals to see how they deal with this issue. These were the July 2010 issues of the American Journal of Pathology and Brain Pathology and the August 2010 issue of the Journal of Neuropathology and Experimental Neurology. In our view, these are mainstream journals and reflect the contemporary state of science in experimental and clinical pathology and neuropathology. Even this brief excursion into the literature revealed many papers in which two-dimensional histology was used to generate numerical data. Many of these studies used software programs designed to collect data, as we have done. A selected sample of these papers include determination of the counts, density, and/or percentage of the following: BrdUrd-positive nuclei in mouse uterus (Buhimschi et al. 2010), cells expressing e-cadherin or fibroblast specific protein in gingival epithelium (Sume et al. 2010), malignant and benign prostatic cells in tissue grafts (Zhao et al. 2010), the numbers of astrocytes and microglia in cerebral cortex and hippocampus of amyloid precursor protein transgenic mice (Huttunen et al. 2010), immunoreactive oligodendrocytes in subcortical white matter of rats exposed to nitrous oxide (Olivier et al. 2010), IL-6 immunoreactive altered neurons in cortical dysplasias (Orlova et al. 2010), and MAP-2 positive neurons on cortical regions in periventricular leukomalacia (Andiman et al. 2010). None of these studies employed stereology. Thus, journals continue to accept cell count data derived from examination of “two-dimensional” tissue sections as a valid morphometric approach. Such an editorial policy is supported in a review by Benes and Lange (2010), who indicate that two-dimensional and three-dimensional approaches to cell counting each have relative strengths and weaknesses, and both are capable of providing estimates of number and numerical density of objects. While we respect stereology as an important quantitative method, we cannot agree with Dr. Butt’s absolute rejection of other morphometric approaches in histopathology in general and for our paper in particular. Relative to the latter, there is another issue we wish to raise. In our study, 1 µm thick sections of archived blocks of tissue embedded in epoxy resin were employed for light microscopic analyses. Sections of this thickness provide optimal resolution at the light microscope level and enabled us to recognize small vacuoles, as shown in Figure 1A of our paper. The stereological probe most commonly used for cell counting is the optical fractionator, which requires the evaluation of 20 to 30 µm thick sections to allow counts to be made in the z-axis of the section. However, the diminished resolution inherent in such thick sections might have obscured the smallest vacuoles and, thus, would have provided a different data set with the possibility of differing interpretation of the available data.
We appreciate Dr. Butt’s interest in our paper, the issues he has raised, and the opportunity to respond to his concerns.