| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Correspondence |
University of Bern, Bern, Switzerland
Karolinska Institut, Stockholm, Sweden
To the Editor-in-Chief:
We contact you in regard to the recent publication in your journal of Rissanen et al1 on the apparent up-regulation of VEGF protein in regenerating fibers in human ischemic skeletal muscle.
Some of the conclusions drawn by the authors appear not to be well founded. Most importantly, we think that the authors have to show critical antibody controls for the VEGF staining. This is essential to exclude the possibility of non-VEGF-specific staining of regenerating fibers in ischemic muscles.
Another main concern relates to the immunohistochemical data. We can demonstrate with a similar detection system that inclusion of any polyclonal first antibody from rabbit in the immunostaining procedure will cause staining of the cytoplasm of "certain" fibers in regenerating but not normal and atrophied (rat) soleus muscle. We have detected in 14 days regenerating muscle exactly the same pattern of immunoreactive fibers as in the study by Rissanen et al1
when we used polyclonal VEGF antiserum or normal rabbit serum (but not PBS alone) as first antibody. According to morphological criteria, the "polyclonal first antiserum" positive fibers in this experiment were found to represent small (regenerating) fibers. A representative example of this experiment is shown in Figure 1
. As far as we know, there is no clear cut molecular explanation for this phenomenon but we guess that the muscle regeneration-specific fiber staining may reflect accumulation of IgG, IgG-binding protein (of the complement system) in damaged fibers. A representative example of this experiment is shown in Figure 1
.
|
of their paper also does not provide the information necessary to exclude potential VEGF-unspecific fiber staining has occurred as they do not derive from consecutive sections to the VEGF protein data. If a similar immune-reaction occurs in regenerating human muscle, the immunoreactive fibers in the ischemic human legs may arise in consequence of inclusion of a polyclonal first antibody, and not due to its specificity for VEGF, in the immunostaining procedure. This may explain Rissanen’s1
conclusions of VEGF’s association with regenerating fibers and macrophages.
To judge the validity of the conclusions of VEGF stain presented in Figure 1
of the Rissanen paper it is therefore important to demonstrate that no fiber-internal staining is observed with normal serum on sections of ischemic muscle consecutive to the ones stained for VEGF.
As a secondary point, some concerns arise due to the lack of quantitative data on the association of VEGF with regenerating and atrophic fibers as well as capillaries. From the description of the authors it is not clear if VEGF could be found in all atrophic and regenerated fibers or if other possible relations as for example to fiber-types existed. Also, the authors do not state how the counting of capillaries were performed, nor do they state the different frequencies in capillary number in VEGF-positive and -negative fibers and its statistical significance. However, they put forward in their conclusions that the amount of capillaries was increased around the atrophic muscle fibers. In this manner the authors do not acknowledge the fact that a clear description of how this measure was done (capillary density, capillary per fiber, fiber per capillary) is of major importance to understanding the vascular adaptations.2
Moreover, to us it is not clear if the strongly VEGF positive lumen in the capillary marker CD31 positive structures reflect thrombus formation in blood vessels and if these would contain VEGF (compare Figure 1, e and f
). In view of the practical implications we felt it to be necessary to comment on this issue. We would very much appreciate to hear other opinions on these issues.
References
University of Kuopio, Kuopio, Finland
Authors’ Reply:
We thank Drs. Gustafsson and Flück for their interest in our study. They express concerns about the specificity of VEGF immunostainings done with rabbit polyclonal antiserum on rat skeletal muscle. However, it is difficult to see how these comments relate to our paper since we have used neither rabbit polyclonal antibodies nor rat tissues.1 All VEGF immunostainings were done on human and rabbit tissues with a mouse monoclonal antibody (Santa Cruz, Clone cs-7269). Also, unlike Drs. Gustafsson and Flück mention in their letter, all immunostainings were controlled by class- and species-matched immunoglobulins in addition to controls where the primary antibody was omitted. All these points are clearly written in the Methods section. In addition, normal horse serum that was used to block possible unspecific staining showed no staining. VEGF immunostainings were also confirmed by in situ hybridizations, which unequivocally demonstrate synthesis of VEGF mRNA in atrophic and regenerating myocytes. It is surprising that Drs. Gustafsson and Flück do not recognize the importance of these findings.
Our results are in agreement with VEGF expression in rat myocardial infarction where the acute widespread VEGF expression was later replaced by VEGF expression restricted to the regions bordering to the infarcted area.2 Also, in a similar pattern with VEGF expression, regenerating but not normal skeletal myocytes have been previously found to express hepatocyte growth factor3 and insulin-like growth factors4 both of which can up-regulate VEGF expression.
Drs. Gustafsson and Flück do not provide any clear explanation for their own findings, which are shown in Figure 1
of their letter, although they speculate possible accumulation of IgG in damaged fibers. In our lab where we have over 20 years’ experience in immunostainings and elsewhere it is well known that immunoglobulins can bind to cytoplasm of dying or damaged cells.5
For that reason, special care was taken to include adequate controls for the immunostainings by Rissanen et al1
Also, the very specific localized immunostaining for VEGF shown in Figure 4i of our paper1
is difficult to discount based on the points raised by Drs. Gustafsson and Flück.
Regarding the comment about the lack of quantitative data, it is generally accepted that it is very difficult to provide quantitative immunohistological data. We used a semi-quantitative grading system for VEGF expression that is applicable to this type of studies. The statements on the increased vasculature in VEGF-positive areas in our paper are qualitative and based on consecutive CD31 and VEGF immunostainings. It is also noteworthy that capillary density and capillary/myocyte ratio may not always be the best measures for VEGF-driven angiogenesis in skeletal muscle.6,7
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
