Immunoelectron Microscopic Characterization of Vasopressin-Producing Neurons in the Hypothalamo-Pituitary Axis of Non-Human Primates by Use of Formaldehyde-Fixed Tissues Stored at −25 °C for Several Years

  1. Akito Otubo
  2. Sho Maejima
  3. Takumi Oti
  4. Keita Satoh
  5. Yasumasa Ueda
  6. John F. Morris
  7. Tatsuya Sakamoto
  8. Hirotaka Sakamoto

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This manuscript shows that it is possible to detect high-abundance peptide antigens in nerve cells at the electron microscope (EM) level in sections of formaldehyde-fixed monkey brain after the sections have been stored for several years in an antifreeze solution in the freezer. The topic of utilizing formalin fixed tissue for research, especially with the numerous "brain banks" worldwide, is an important topic especially if one wishes to conduct studies in post mortem human tissue. The authors used antibodies to detect the presence of vasopressin gene-related products (i.e., neurophysin II and copeptin) in the hypothalamus and pituitary of the monkey brain. This paper is of interest to anatomists who work on AVP neurons in non-human primate. Due to issues with tissue quality, methodology and interpretation, the experimental approach described in this paper may not be as useful for studying fixed and archived brain sections as the authors conclude.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #3 agreed to share their name with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Translational research often requires the testing of experimental therapies in primates, but research in non-human primates is now stringently controlled by law around the world. Tissues fixed in formaldehyde without glutaraldehyde have been thought to be inappropriate for use in electron microscopic analysis, particularly those of the brain. Here we report the immunoelectron microscopic characterization of arginine vasopressin (AVP)-producing neurons in macaque hypothalamo-pituitary axis tissues fixed by perfusion with 4% formaldehyde and stored at −25 °C for several years (4–6 years). The size difference of dense-cored vesicles between magnocellular and parvocellular AVP neurons was detectable in their cell bodies and perivascular nerve endings located, respectively, in the posterior pituitary and median eminence. Furthermore, glutamate and the vesicular glutamate transporter 2 could be colocalized with AVP in perivascular nerve endings of both the posterior pituitary and the external layer of the median eminence, suggesting that both magnocellular and parvocellular AVP neurons are glutamatergic in primates. Both ultrastructure and immunoreactivity can therefore be sufficiently preserved in macaque brain tissues stored long-term, initially for light microscopy. Taken together, these results suggest that this methodology could be applied to the human post-mortem brain and be very useful in translational research.

Article activity feed

  1. eLife

    Evaluation Summary:

    This manuscript shows that it is possible to detect high-abundance peptide antigens in nerve cells at the electron microscope (EM) level in sections of formaldehyde-fixed monkey brain after the sections have been stored for several years in an antifreeze solution in the freezer. The topic of utilizing formalin fixed tissue for research, especially with the numerous "brain banks" worldwide, is an important topic especially if one wishes to conduct studies in post mortem human tissue. The authors used antibodies to detect the presence of vasopressin gene-related products (i.e., neurophysin II and copeptin) in the hypothalamus and pituitary of the monkey brain. This paper is of interest to anatomists who work on AVP neurons in non-human primate. Due to issues with tissue quality, methodology and interpretation, the experimental approach described in this paper may not be as useful for studying fixed and archived brain sections as the authors conclude.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #3 agreed to share their name with the authors.)

  2. eLife

    Reviewer #1 (Public Review):

    This study conducted by Akito Otubo et. al has two goals: 1. evaluate the usefulness of using formalin fixed frozen tissue that has been stored for several years, and 2. characterization of arginine vasopressin (AVP)-producing neurons in macaque hypothalamo-pituitary axis using immunoelectron microscopy approaches. The authors seek to follow mouse studies that suggest co-release of glutamate and corticotrophin-releasing factor (CRF) by magnocellular neurosecretory neurons in the supraoptic nucleus (SON) and paraventricular nucleus (PVN) of the hypothalamus. The specific goal being to ask if a similar co-release mechanism occurs in the primate AVP/CRF system.

    The major strength of the results is that they do show antigenicity in formalin fixed tissue, but the major weaknesses listed below leave me unconvinced by their conclusions that, "We found that both ultrastructure and immunoreactivity are sufficiently preserved in macaque brain tissues stored long-term for light microscopy", and thus I do not believe they have achieved their aims. There are three main issues I have: 1. The quality of the tissue is extremely poor as there are numerous membrane breaks making it near impossible to make out cellular structures. For instance, without the antibody staining to guide the eye, I question whether any cellular structures could be made out, 2. it's not stated whether the antibodies used in this study were the ones that just happened to work or if antibodies work universally; such burden of proof is essential if the authors wish to claim that old formalin fixed tissue is of value, and 3. there's a significant lack of controls on two fronts: a. controls with a properly fixed (fresh, glutaraldedhyde, etc.) brain showing antigenicity is similar to the old formalin tissue, and b. negative controls for the co-release model that prove the immunostaining is specific. For example, staining for a protein that shouldn't localize in the PCV (and hence not co-localize with NPII or copeptin). There lacks similar negative controls for the immunofluorescence data. The burden of extensive controls is on these authors if they wish to establish that older tissue is of scientific value. Overall, without controls showing 1. Old formalin fixed tissue compared to fresh tissue show equivalent results, 2. antigenicity is in fact real, and 3. antigenicity is broadly true for several biological markers.

  3. eLife

    Reviewer #2 (Public Review):

    Otubo et al aimed to use immunoelectron microscopic method to detect arginine vasopressin (AVP)-expressing neurons in the hypothalamo-pituitary axis of macaque brain stored in 10% formalin. They performed (1) Western blotting to confirm the specificity of the antibodies, (2) immunofluorescent staining to vasopressin-associated neurophysin (NPII) in the macaque hypothalamus, and (3) double-label immunoelectron microscopy and detected NPII and copeptin in the posterior pituitary and the median eminence. They also (4) quantified the size differences of dcv in different hypothalamic sub-regions.

    The main finding was the smaller dcv size in external median eminence (when compared to posterior pituitary and internal eminence). They also found 2 subgroups of AVP neurons in the PVN based on dcv size (the larger magnocellular and smaller parvocellular). They also confirmed previous finding that a subpopulation of the CRF co-expressing parvocellular neurons were more intermingled with magnocellular neurons in monkey than in mice, and the glutamatergic identities of the magno- and parvocellular AVP neurons.

    The strength of the paper is that it quantified the size differences of dcv between different sub-regions of the hypothalamus. However, the author claimed that formaldehyde-fixed materials without glutaraldehyde have long been thought to be inappropriate for electron microscopic observation. However, formalin-fixed tissue has been shown to be more than adequate for quality electron microscopic analysis (PMIDs: 17947985, 29867279).

    The conclusions of this paper are mostly well supported by data, with the exception of the suitability of formalin-fixed tissue for electron microscopic analysis. The experiments confirmed previous knowledge but in its current form this manuscript does not provide additional insights or technical advances.

  4. eLife

    Reviewer #3 (Public Review):

    In this study, Otubo and colleagues describe the detection of synaptic vesicle-associated peptide antigens, glutamate and VGLUT2 at the electron microscope (EM) level in hypothalamus and pituitary from macaque brains that were perfused with paraformaldehyde, cryostat-sectioned and stored in cryoprotectant at -25oC for up to 5 years. Formaldehyde-fixed tissue is generally considered unsuitable for any type of EM study so Otubo et al's success with EM immunostaining is surprising. Congratulations to them for ignoring conventional wisdom and attempting EM immunolabelling on sections of formaldehyde-fixed tissue after long term storage. Unfortunately, issues with the experimental approaches used by the authors mean that a number of their conclusions are not soundly based. The authors demonstrate that their EM immunogold staining protocol is good for detecting high-abundance antigens, like peptides and proteins that are stored in dense cored vesicles (DCV); but there is no data on antigens that are present at low concentrations. This possible shortcoming could render Obuto and colleague's protocol of limited use for future studies on fixed and archived human and primate brain samples.

    STRENGTHS

    The immunofluorescent staining for vasopressin-associated neurophysin (NPII), copeptin and corticotropin releasing factor (CRF) shown in Figures 2, 8 and 9 in this manuscript is beautiful and confirms the quality of results that can be obtained from sections stored long term in cryoprotectant at -25oC. Nevertheless, this finding is not surprising or novel. Banking sections of formaldehyde-fixed tissue in anti-freeze solution in the freezer for future immunohistochemical processing is common practice in many laboratories around the world. For example, in my laboratory, we have immunohistochemically processed brain sections stored in this way for up to 20 years and obtained staining that is as good as in freshly-cut sections from the same brains.

    Otubo et al's successful localization of NPII, copeptin and CRF at the EM level is obvious from their micrographs showing post-embedding immunogold staining. The choice of these antigens was a good one. Peptides/proteins, like vasopressin (AVP), NPII, copeptin and CRF, are highly concentrated in the DCV of peptidergic neurons. As this study shows, even if considerable antigen is lost due to suboptimal primary fixation, enough persists in DCV to be localized with the method described in this manuscript. It would be interesting to know whether less abundant antigens, such as enzymes or receptors, can be localized at the EM level in sections of formaldehyde-fixed tissue after long term storage at -25oC. If low abundance antigens can be detected, the authors' protocol would be much more useful.

    WEAKNESSES

    The Results section does not highlight several technical details (documented in the following Methods section) that are vitally important for assessing and interpreting Otubo et al's findings. In Methods, we learn that cryostat sections rather than intact brain blocks were stored at -25oC in a standard cryoprotectant solution for up to 5 years. After removal from cryoprotectant, the sections were post-fixed with glutaraldehyde to improve retention of antigens and then dehydrated and embedded into resin without prior treatment with osmium tetroxide (OsO4).

    Based on these technical details, I question Otubo's et al's assertions that immunoreactivity for any of the antigens investigated is associated with membranes or occurs within membrane-bound structures. Aldehyde fixatives preserve peptides, proteins and nucleic acids. OsO4 treatment is required for the preservation of lipids in membranes. Without this membrane preservation step, lipids are dissolve away by the solvents used for dehydrating sections (i.e., alcohols, acetone and propylene oxide) before the sections are embedded in resin. Because Otubo et al omitted OsO4 and dehydrated their sections through methanol, there cannot be any membranes in the samples that they examined ultrastructurally. They do comment that "the membrane of single microvesicles could not be clearly distinguished" and may have been deceived into thinking that membranes were present. In their ultrathin sections, membrane-bound proteins appear as linear arrays despite extraction of the lipid bilayer with which the proteins were associated. Interpreting ultrastructure without the presence of membranes is a serious problem, which is illustrated by what the authors call MV, i.e., clusters of "microvesicles". In my opinion, without membranes to show the boundaries of cellular and subcellular structures, it is not possible to unequivocally identify synaptic vesicles (i.e., "microvesicles") in axon terminals. For example, the indistinct EM appearance of the lower left MV in Figure 4C could just as likely be Golgi apparatus as clustered synaptic vesicles.

    The authors' demonstration of neurotransmitter glutamate in macaque AVP neurons is suspect. All cells use glutamate to make protein; in addition, some neurons use glutamate as a neurotransmitter. Before claiming that a neuron is glutamatergic based on immunogold labelling, it is mandatory to establish that the density of gold particles over axon terminals is significantly different from the background density (see, for example, Llewellyn-Smith et al, 1992, 1997, 1998). Unless the authors perform this type of quantitative analysis of gold particle densities, their claim that their EM immunostaining has revealed glutamate in macaque vasopressin (AVP) neurons is not justified.

    The immunogold labelling pattern that the authors achieved with anti-VGLUT2 is perplexing. VGLUT2 is involved in packaging glutamate into synaptic vesicles and these glutamate-containing vesicles are generally small and clear. Thus, I would expect that small gold particles detecting VGLUT2-immunoreactivity would be concentrated over MV, i.e., clumps of "microvesicles". In Figure 13, however, small gold particles are sparse over MV, where labelling would be expected, and appear in similar density on/near DCV, where the concentration of neurotransmitter glutamate is expected to be low. The observed distribution of gold particles suggests that the antibody may be detecting something other than VGLUT2. Although the anti-VGLUT2 used in this study shows a single band on western blotting, fixation and tissue processing can alter epitopes on antigens. It would be advisable for the authors to confirm the antibody's specificity on their tissue by doing immunogold staining with anti-VGLUT2 absorbed with the antigen against which it was raised.

    Otubo et al make a number of conclusions from their measurements of vesicle sizes. Figure 5 contains enough data to confirm statistically that DCV in the external median eminence (ME) are smaller than DCV in either the internal ME or the posterior pituitary - as long as a reasonable numbers of DCV were measured in sections from each of at least 3 of the 4 monkey used for anatomical studies. An unknown number of monkeys also provided the data presented in Table 2.

    CONCLUSIONS

    As indicated by its title, the most interesting aspect of this paper is the use of years-old sections of formaldehyde-fixed tissue for EM localization of neuropeptides in brain tissue. The applicability of this method for studying banked sections of human and primate brains will depend upon whether or not low abundance antigens can be detected using the methods described here.

    The paper provides only an incremental advance in our understanding of AVP circuitry in the monkey hypothalamus and pituitary. Because the results on glutamate and VGLUT2 immunoreactivity in the present study are questionable, the only reliable new information is on the co-existence (as expected) of copeptin, one of the vasopressin gene products, with another vasopressin gene product, NPII, in AVP neurons. Otubo and colleagues have previously published the co-localization of CRF in AVP neurons in their 2020 paper in J Neuroendocrinol.

  5. Version published to 10.3390/ijms22179180
  6. Version published to 10.1101/2021.06.07.447309v1 on bioRxiv